Entries Tagged "democracy"

Page 5 of 6

AI to Aid Democracy

There’s good reason to fear that AI systems like ChatGPT and GPT4 will harm democracy. Public debate may be overwhelmed by industrial quantities of autogenerated argument. People might fall down political rabbit holes, taken in by superficially convincing bullshit, or obsessed by folies à deux relationships with machine personalities that don’t really exist.

These risks may be the fallout of a world where businesses deploy poorly tested AI systems in a battle for market share, each hoping to establish a monopoly.

But dystopia isn’t the only possible future. AI could advance the public good, not private profit, and bolster democracy instead of undermining it. That would require an AI not under the control of a large tech monopoly, but rather developed by government and available to all citizens. This public option is within reach if we want it.

An AI built for public benefit could be tailor-made for those use cases where technology can best help democracy. It could plausibly educate citizens, help them deliberate together, summarize what they think, and find possible common ground. Politicians might use large language models, or LLMs, like GPT4 to better understand what their citizens want.

Today, state-of-the-art AI systems are controlled by multibillion-dollar tech companies: Google, Meta, and OpenAI in connection with Microsoft. These companies get to decide how we engage with their AIs and what sort of access we have. They can steer and shape those AIs to conform to their corporate interests. That isn’t the world we want. Instead, we want AI options that are both public goods and directed toward public good.

We know that existing LLMs are trained on material gathered from the internet, which can reflect racist bias and hate. Companies attempt to filter these data sets, fine-tune LLMs, and tweak their outputs to remove bias and toxicity. But leaked emails and conversations suggest that they are rushing half-baked products to market in a race to establish their own monopoly.

These companies make decisions with huge consequences for democracy, but little democratic oversight. We don’t hear about political trade-offs they are making. Do LLM-powered chatbots and search engines favor some viewpoints over others? Do they skirt controversial topics completely? Currently, we have to trust companies to tell us the truth about the trade-offs they face.

A public option LLM would provide a vital independent source of information and a testing ground for technological choices with big democratic consequences. This could work much like public option health care plans, which increase access to health services while also providing more transparency into operations in the sector and putting productive pressure on the pricing and features of private products. It would also allow us to figure out the limits of LLMs and direct their applications with those in mind.

We know that LLMs often “hallucinate,” inferring facts that aren’t real. It isn’t clear whether this is an unavoidable flaw of how they work, or whether it can be corrected for. Democracy could be undermined if citizens trust technologies that just make stuff up at random, and the companies trying to sell these technologies can’t be trusted to admit their flaws.

But a public option AI could do more than check technology companies’ honesty. It could test new applications that could support democracy rather than undermining it.

Most obviously, LLMs could help us formulate and express our perspectives and policy positions, making political arguments more cogent and informed, whether in social media, letters to the editor, or comments to rule-making agencies in response to policy proposals. By this we don’t mean that AI will replace humans in the political debate, only that they can help us express ourselves. If you’ve ever used a Hallmark greeting card or signed a petition, you’ve already demonstrated that you’re OK with accepting help to articulate your personal sentiments or political beliefs. AI will make it easier to generate first drafts, and provide editing help and suggest alternative phrasings. How these AI uses are perceived will change over time, and there is still much room for improvement in LLMs—but their assistive power is real. People are already testing and speculating on their potential for speechwriting, lobbying, and campaign messaging. Highly influential people often rely on professional speechwriters and staff to help develop their thoughts, and AI could serve a similar role for everyday citizens.

If the hallucination problem can be solved, LLMs could also become explainers and educators. Imagine citizens being able to query an LLM that has expert-level knowledge of a policy issue, or that has command of the positions of a particular candidate or party. Instead of having to parse bland and evasive statements calibrated for a mass audience, individual citizens could gain real political understanding through question-and-answer sessions with LLMs that could be unfailingly available and endlessly patient in ways that no human could ever be.

Finally, and most ambitiously, AI could help facilitate radical democracy at scale. As Carnegie Mellon professor of statistics Cosma Shalizi has observed, we delegate decisions to elected politicians in part because we don’t have time to deliberate on every issue. But AI could manage massive political conversations in chat rooms, on social networking sites, and elsewhere: identifying common positions and summarizing them, surfacing unusual arguments that seem compelling to those who have heard them, and keeping attacks and insults to a minimum.

AI chatbots could run national electronic town hall meetings and automatically summarize the perspectives of diverse participants. This type of AI-moderated civic debate could also be a dynamic alternative to opinion polling. Politicians turn to opinion surveys to capture snapshots of popular opinion because they can only hear directly from a small number of voters, but want to understand where voters agree or disagree.

Looking further into the future, these technologies could help groups reach consensus and make decisions. Early experiments by the AI company DeepMind suggest that LLMs can build bridges between people who disagree, helping bring them to consensus. Science fiction writer Ruthanna Emrys, in her remarkable novel A Half-Built Garden, imagines how AI might help people have better conversations and make better decisions—rather than taking advantage of these biases to maximize profits.

This future requires an AI public option. Building one, through a government-directed model development and deployment program, would require a lot of effort—and the greatest challenges in developing public AI systems would be political.

Some technological tools are already publicly available. In fairness, tech giants like Google and Meta have made many of their latest and greatest AI tools freely available for years, in cooperation with the academic community. Although OpenAI has not made the source code and trained features of its latest models public, competitors such as Hugging Face have done so for similar systems.

While state-of-the-art LLMs achieve spectacular results, they do so using techniques that are mostly well known and widely used throughout the industry. OpenAI has only revealed limited details of how it trained its latest model, but its major advance over its earlier ChatGPT model is no secret: a multi-modal training process that accepts both image and textual inputs.

Financially, the largest-scale LLMs being trained today cost hundreds of millions of dollars. That’s beyond ordinary people’s reach, but it’s a pittance compared to U.S. federal military spending—and a great bargain for the potential return. While we may not want to expand the scope of existing agencies to accommodate this task, we have our choice of government labs, like the National Institute of Standards and Technology, the Lawrence Livermore National Laboratory, and other Department of Energy labs, as well as universities and nonprofits, with the AI expertise and capability to oversee this effort.

Instead of releasing half-finished AI systems for the public to test, we need to make sure that they are robust before they’re released—and that they strengthen democracy rather than undermine it. The key advance that made recent AI chatbot models dramatically more useful was feedback from real people. Companies employ teams to interact with early versions of their software to teach them which outputs are useful and which are not. These paid users train the models to align to corporate interests, with applications like web search (integrating commercial advertisements) and business productivity assistive software in mind.

To build assistive AI for democracy, we would need to capture human feedback for specific democratic use cases, such as moderating a polarized policy discussion, explaining the nuance of a legal proposal, or articulating one’s perspective within a larger debate. This gives us a path to “align” LLMs with our democratic values: by having models generate answers to questions, make mistakes, and learn from the responses of human users, without having these mistakes damage users and the public arena.

Capturing that kind of user interaction and feedback within a political environment suspicious of both AI and technology generally will be challenging. It’s easy to imagine the same politicians who rail against the untrustworthiness of companies like Meta getting far more riled up by the idea of government having a role in technology development.

As Karl Popper, the great theorist of the open society, argued, we shouldn’t try to solve complex problems with grand hubristic plans. Instead, we should apply AI through piecemeal democratic engineering, carefully determining what works and what does not. The best way forward is to start small, applying these technologies to local decisions with more constrained stakeholder groups and smaller impacts.

The next generation of AI experimentation should happen in the laboratories of democracy: states and municipalities. Online town halls to discuss local participatory budgeting proposals could be an easy first step. Commercially available and open-source LLMs could bootstrap this process and build momentum toward federal investment in a public AI option.

Even with these approaches, building and fielding a democratic AI option will be messy and hard. But the alternative—shrugging our shoulders as a fight for commercial AI domination undermines democratic politics—will be much messier and much worse.

This essay was written with Henry Farrell and Nathan Sanders, and previously appeared on Slate.com.

EDITED TO ADD: Linux Weekly News discussion.

EDITED TO ADD: This post has been translated into Hebrew.

Posted on April 26, 2023 at 6:51 AM • View Comments

How AI Could Write Our Laws

Nearly 90% of the multibillion-dollar federal lobbying apparatus in the United States serves corporate interests. In some cases, the objective of that money is obvious. Google pours millions into lobbying on bills related to antitrust regulation. Big energy companies expect action whenever there is a move to end drilling leases for federal lands, in exchange for the tens of millions they contribute to congressional reelection campaigns.

But lobbying strategies are not always so blunt, and the interests involved are not always so obvious. Consider, for example, a 2013 Massachusetts bill that tried to restrict the commercial use of data collected from K-12 students using services accessed via the internet. The bill appealed to many privacy-conscious education advocates, and appropriately so. But behind the justification of protecting students lay a market-altering policy: the bill was introduced at the behest of Microsoft lobbyists, in an effort to exclude Google Docs from classrooms.

What would happen if such legal-but-sneaky strategies for tilting the rules in favor of one group over another become more widespread and effective? We can see hints of an answer in the remarkable pace at which artificial-intelligence tools for everything from writing to graphic design are being developed and improved. And the unavoidable conclusion is that AI will make lobbying more guileful, and perhaps more successful.

It turns out there is a natural opening for this technology: microlegislation.

“Microlegislation” is a term for small pieces of proposed law that cater—sometimes unexpectedly—to narrow interests. Political scientist Amy McKay coined the term. She studied the 564 amendments to the Affordable Care Act (“Obamacare”) considered by the Senate Finance Committee in 2009, as well as the positions of 866 lobbying groups and their campaign contributions. She documented instances where lobbyist comments—on health-care research, vaccine services, and other provisions—were translated directly into microlegislation in the form of amendments. And she found that those groups’ financial contributions to specific senators on the committee increased the amendments’ chances of passing.

Her finding that lobbying works was no surprise. More important, McKay’s work demonstrated that computer models can predict the likely fate of proposed legislative amendments, as well as the paths by which lobbyists can most effectively secure their desired outcomes. And that turns out to be a critical piece of creating an AI lobbyist.

Lobbying has long been part of the give-and-take among human policymakers and advocates working to balance their competing interests. The danger of microlegislation—a danger greatly exacerbated by AI—is that it can be used in a way that makes it difficult to figure out who the legislation truly benefits.

Another word for a strategy like this is a “hack.” Hacks follow the rules of a system but subvert their intent. Hacking is often associated with computer systems, but the concept is also applicable to social systems like financial markets, tax codes, and legislative processes.

While the idea of monied interests incorporating AI assistive technologies into their lobbying remains hypothetical, specific machine-learning technologies exist today that would enable them to do so. We should expect these techniques to get better and their utilization to grow, just as we’ve seen in so many other domains.

Here’s how it might work.

Crafting an AI microlegislator

To make microlegislation, machine-learning systems must be able to uncover the smallest modification that could be made to a bill or existing law that would make the biggest impact on a narrow interest.

There are three basic challenges involved. First, you must create a policy proposal—small suggested changes to legal text—and anticipate whether or not a human reader would recognize the alteration as substantive. This is important; a change that isn’t detectable is more likely to pass without controversy. Second, you need to do an impact assessment to project the implications of that change for the short- or long-range financial interests of companies. Third, you need a lobbying strategizer to identify what levers of power to pull to get the best proposal into law.

Existing AI tools can tackle all three of these.

The first step, the policy proposal, leverages the core function of generative AI. Large language models, the sort that have been used for general-purpose chatbots such as ChatGPT, can easily be adapted to write like a native in different specialized domains after seeing a relatively small number of examples. This process is called fine-tuning. For example, a model “pre-trained” on a large library of generic text samples from books and the internet can be “fine-tuned” to work effectively on medical literature, computer science papers, and product reviews.

Given this flexibility and capacity for adaptation, a large language model could be fine-tuned to produce draft legislative texts, given a data set of previously offered amendments and the bills they were associated with. Training data is available. At the federal level, it’s provided by the US Government Publishing Office, and there are already tools for downloading and interacting with it. Most other jurisdictions provide similar data feeds, and there are even convenient assemblages of that data.

Meanwhile, large language models like the one underlying ChatGPT are routinely used for summarizing long, complex documents (even laws and computer code) to capture the essential points, and they are optimized to match human expectations. This capability could allow an AI assistant to automatically predict how detectable the true effect of a policy insertion may be to a human reader.

Today, it can take a highly paid team of human lobbyists days or weeks to generate and analyze alternative pieces of microlegislation on behalf of a client. With AI assistance, that could be done instantaneously and cheaply. This opens the door to dramatic increases in the scope of this kind of microlegislating, with a potential to scale across any number of bills in any jurisdiction.

Teaching machines to assess impact

Impact assessment is more complicated. There is a rich series of methods for quantifying the predicted outcome of a decision or policy, and then also optimizing the return under that model. This kind of approach goes by different names in different circles—mathematical programming in management science, utility maximization in economics, and rational design in the life sciences.

To train an AI to do this, we would need to specify some way to calculate the benefit to different parties as a result of a policy choice. That could mean estimating the financial return to different companies under a few different scenarios of taxation or regulation. Economists are skilled at building risk models like this, and companies are already required to formulate and disclose regulatory compliance risk factors to investors. Such a mathematical model could translate directly into a reward function, a grading system that could provide feedback for the model used to create policy proposals and direct the process of training it.

The real challenge in impact assessment for generative AI models would be to parse the textual output of a model like ChatGPT in terms that an economic model could readily use. Automating this would require extracting structured financial information from the draft amendment or any legalese surrounding it. This kind of information extraction, too, is an area where AI has a long history; for example, AI systems have been trained to recognize clinical details in doctors’ notes. Early indications are that large language models are fairly good at recognizing financial information in texts such as investor call transcripts. While it remains an open challenge in the field, they may even be capable of writing out multi-step plans based on descriptions in free text.

Machines as strategists

The last piece of the puzzle is a lobbying strategizer to figure out what actions to take to convince lawmakers to adopt the amendment.

Passing legislation requires a keen understanding of the complex interrelated networks of legislative offices, outside groups, executive agencies, and other stakeholders vying to serve their own interests. Each actor in this network has a baseline perspective and different factors that influence that point of view. For example, a legislator may be moved by seeing an allied stakeholder take a firm position, or by a negative news story, or by a campaign contribution.

It turns out that AI developers are very experienced at modeling these kinds of networks. Machine-learning models for network graphs have been built, refined, improved, and iterated by hundreds of researchers working on incredibly diverse problems: lidar scans used to guide self-driving cars, the chemical functions of molecular structures, the capture of motion in actors’ joints for computer graphics, behaviors in social networks, and more.

In the context of AI-assisted lobbying, political actors like legislators and lobbyists are nodes on a graph, just like users in a social network. Relations between them are graph edges, like social connections. Information can be passed along those edges, like messages sent to a friend or campaign contributions made to a member. AI models can use past examples to learn to estimate how that information changes the network. Calculating the likelihood that a campaign contribution of a given size will flip a legislator’s vote on an amendment is one application.

McKay’s work has already shown us that there are significant, predictable relationships between these actions and the outcomes of legislation, and that the work of discovering those can be automated. Others have shown that graphs of neural network models like those described above can be applied to political systems. The full-scale use of these technologies to guide lobbying strategy is theoretical, but plausible.

Put together, these three components could create an automatic system for generating profitable microlegislation. The policy proposal system would create millions, even billions, of possible amendments. The impact assessor would identify the few that promise to be most profitable to the client. And the lobbying strategy tool would produce a blueprint for getting them passed.

What remains is for human lobbyists to walk the floors of the Capitol or state house, and perhaps supply some cash to grease the wheels. These final two aspects of lobbying—access and financing—cannot be supplied by the AI tools we envision. This suggests that lobbying will continue to primarily benefit those who are already influential and wealthy, and AI assistance will amplify their existing advantages.

The transformative benefit that AI offers to lobbyists and their clients is scale. While individual lobbyists tend to focus on the federal level or a single state, with AI assistance they could more easily infiltrate a large number of state-level (or even local-level) law-making bodies and elections. At that level, where the average cost of a seat is measured in the tens of thousands of dollars instead of millions, a single donor can wield a lot of influence—if automation makes it possible to coordinate lobbying across districts.

How to stop them

When it comes to combating the potentially adverse effects of assistive AI, the first response always seems to be to try to detect whether or not content was AI-generated. We could imagine a defensive AI that detects anomalous lobbyist spending associated with amendments that benefit the contributing group. But by then, the damage might already be done.

In general, methods for detecting the work of AI tend not to keep pace with its ability to generate convincing content. And these strategies won’t be implemented by AIs alone. The lobbyists will still be humans who take the results of an AI microlegislator and further refine the computer’s strategies. These hybrid human-AI systems will not be detectable from their output.

But the good news is: the same strategies that have long been used to combat misbehavior by human lobbyists can still be effective when those lobbyists get an AI assist. We don’t need to reinvent our democracy to stave off the worst risks of AI; we just need to more fully implement long-standing ideals.

First, we should reduce the dependence of legislatures on monolithic, multi-thousand-page omnibus bills voted on under deadline. This style of legislating exploded in the 1980s and 1990s and continues through to the most recent federal budget bill. Notwithstanding their legitimate benefits to the political system, omnibus bills present an obvious and proven vehicle for inserting unnoticed provisions that may later surprise the same legislators who approved them.

The issue is not that individual legislators need more time to read and understand each bill (that isn’t realistic or even necessary). It’s that omnibus bills must pass. There is an imperative to pass a federal budget bill, and so the capacity to push back on individual provisions that may seem deleterious (or just impertinent) to any particular group is small. Bills that are too big to fail are ripe for hacking by microlegislation.

Moreover, the incentive for legislators to introduce microlegislation catering to a narrow interest is greater if the threat of exposure is lower. To strengthen the threat of exposure for misbehaving legislative sponsors, bills should focus more tightly on individual substantive areas and, after the introduction of amendments, allow more time before the committee and floor votes. During this time, we should encourage public review and testimony to provide greater oversight.

Second, we should strengthen disclosure requirements on lobbyists, whether they’re entirely human or AI-assisted. State laws regarding lobbying disclosure are a hodgepodge. North Dakota, for example, only requires lobbying reports to be filed annually, so that by the time a disclosure is made, the policy is likely already decided. A lobbying disclosure scorecard created by Open Secrets, a group researching the influence of money in US politics, tracks nine states that do not even require lobbyists to report their compensation.

Ideally, it would be great for the public to see all communication between lobbyists and legislators, whether it takes the form of a proposed amendment or not. Absent that, let’s give the public the benefit of reviewing what lobbyists are lobbying for—and why. Lobbying is traditionally an activity that happens behind closed doors. Right now, many states reinforce that: they actually exempt testimony delivered publicly to a legislature from being reported as lobbying.

In those jurisdictions, if you reveal your position to the public, you’re no longer lobbying. Let’s do the inverse: require lobbyists to reveal their positions on issues. Some jurisdictions already require a statement of position (a ‘yea’ or ‘nay’) from registered lobbyists. And in most (but not all) states, you could make a public records request regarding meetings held with a state legislator and hope to get something substantive back. But we can expect more—lobbyists could be required to proactively publish, within a few days, a brief summary of what they demanded of policymakers during meetings and why they believe it’s in the general interest.

We can’t rely on corporations to be forthcoming and wholly honest about the reasons behind their lobbying positions. But having them on the record about their intentions would at least provide a baseline for accountability.

Finally, consider the role AI assistive technologies may have on lobbying firms themselves and the labor market for lobbyists. Many observers are rightfully concerned about the possibility of AI replacing or devaluing the human labor it automates. If the automating potential of AI ends up commodifying the work of political strategizing and message development, it may indeed put some professionals on K Street out of work.

But don’t expect that to disrupt the careers of the most astronomically compensated lobbyists: former members Congress and other insiders who have passed through the revolving door. There is no shortage of reform ideas for limiting the ability of government officials turned lobbyists to sell access to their colleagues still in government, and they should be adopted and—equally important—maintained and enforced in successive Congresses and administrations.

None of these solutions are really original, specific to the threats posed by AI, or even predominantly focused on microlegislation—and that’s the point. Good governance should and can be robust to threats from a variety of techniques and actors.

But what makes the risks posed by AI especially pressing now is how fast the field is developing. We expect the scale, strategies, and effectiveness of humans engaged in lobbying to evolve over years and decades. Advancements in AI, meanwhile, seem to be making impressive breakthroughs at a much faster pace—and it’s still accelerating.

The legislative process is a constant struggle between parties trying to control the rules of our society as they are updated, rewritten, and expanded at the federal, state, and local levels. Lobbying is an important tool for balancing various interests through our system. If it’s well-regulated, perhaps lobbying can support policymakers in making equitable decisions on behalf of us all.

This article was co-written with Nathan E. Sanders and originally appeared in MIT Technology Review.

Posted on March 14, 2023 at 12:01 PM • View Comments

Defending against AI Lobbyists

When is it time to start worrying about artificial intelligence interfering in our democracy? Maybe when an AI writes a letter to The New York Times opposing the regulation of its own technology.

That happened last month. And because the letter was responding to an essay we wrote, we’re starting to get worried. And while the technology can be regulated, the real solution lies in recognizing that the problem is human actors—and those we can do something about.

Our essay argued that the much heralded launch of the AI chatbot ChatGPT, a system that can generate text realistic enough to appear to be written by a human, poses significant threats to democratic processes. The ability to produce high quality political messaging quickly and at scale, if combined with AI-assisted capabilities to strategically target those messages to policymakers and the public, could become a powerful accelerant of an already sprawling and poorly constrained force in modern democratic life: lobbying.

We speculated that AI-assisted lobbyists could use generative models to write op-eds and regulatory comments supporting a position, identify members of Congress who wield the most influence over pending legislation, use network pattern identification to discover undisclosed or illegal political coordination, or use supervised machine learning to calibrate the optimal contribution needed to sway the vote of a legislative committee member.

These are all examples of what we call AI hacking. Hacks are strategies that follow the rules of a system, but subvert its intent. Currently a human creative process, future AIs could discover, develop, and execute these same strategies.

While some of these activities are the longtime domain of human lobbyists, AI tools applied against the same task would have unfair advantages. They can scale their activity effortlessly across every state in the country—human lobbyists tend to focus on a single state—they may uncover patterns and approaches unintuitive and unrecognizable by human experts, and do so nearly instantaneously with little chance for human decision makers to keep up.

These factors could make AI hacking of the democratic process fundamentally ungovernable. Any policy response to limit the impact of AI hacking on political systems would be critically vulnerable to subversion or control by an AI hacker. If AI hackers achieve unchecked influence over legislative processes, they could dictate the rules of our society: including the rules that govern AI.

We admit that this seemed far fetched when we first wrote about it in 2021. But now that the emanations and policy prescriptions of ChatGPT have been given an audience in the New York Times and innumerable other outlets in recent weeks, it’s getting harder to dismiss.

At least one group of researchers is already testing AI techniques to automatically find and advocate for bills that benefit a particular interest. And one Massachusetts representative used ChatGPT to draft legislation regulating AI.

The AI technology of two years ago seems quaint by the standards of ChatGPT. What will the technology of 2025 seem like if we could glimpse it today? To us there is no question that now is the time to act.

First, let’s dispense with the concepts that won’t work. We cannot solely rely on explicit regulation of AI technology development, distribution, or use. Regulation is essential, but it would be vastly insufficient. The rate of AI technology development, and the speed at which AI hackers might discover damaging strategies, already outpaces policy development, enactment, and enforcement.

Moreover, we cannot rely on detection of AI actors. The latest research suggests that AI models trying to classify text samples as human- or AI-generated have limited precision, and are ill equipped to handle real world scenarios. These reactive, defensive techniques will fail because the rate of advancement of the “offensive” generative AI is so astounding.

Additionally, we risk a dragnet that will exclude masses of human constituents that will use AI to help them express their thoughts, or machine translation tools to help them communicate. If a written opinion or strategy conforms to the intent of a real person, it should not matter if they enlisted the help of an AI (or a human assistant) to write it.

Most importantly, we should avoid the classic trap of societies wrenched by the rapid pace of change: privileging the status quo. Slowing down may seem like the natural response to a threat whose primary attribute is speed. Ideas like increasing requirements for human identity verification, aggressive detection regimes for AI-generated messages, and elongation of the legislative or regulatory process would all play into this fallacy. While each of these solutions may have some value independently, they do nothing to make the already powerful actors less powerful.

Finally, it won’t work to try to starve the beast. Large language models like ChatGPT have a voracious appetite for data. They are trained on past examples of the kinds of content that they will be asked to generate in the future. Similarly, an AI system built to hack political systems will rely on data that documents the workings of those systems, such as messages between constituents and legislators, floor speeches, chamber and committee voting results, contribution records, lobbying relationship disclosures, and drafts of and amendments to legislative text. The steady advancement towards the digitization and publication of this information that many jurisdictions have made is positive. The threat of AI hacking should not dampen or slow progress on transparency in public policymaking.

Okay, so what will help?

First, recognize that the true threats here are malicious human actors. Systems like ChatGPT and our still-hypothetical political-strategy AI are still far from artificial general intelligences. They do not think. They do not have free will. They are just tools directed by people, much like lobbyist for hire. And, like lobbyists, they will be available primarily to the richest individuals, groups, and their interests.

However, we can use the same tools that would be effective in controlling human political influence to curb AI hackers. These tools will be familiar to any follower of the last few decades of U.S. political history.

Campaign finance reforms such as contribution limits, particularly when applied to political action committees of all types as well as to candidate operated campaigns, can reduce the dependence of politicians on contributions from private interests. The unfair advantage of a malicious actor using AI lobbying tools is at least somewhat mitigated if a political target’s entire career is not already focused on cultivating a concentrated set of major donors.

Transparency also helps. We can expand mandatory disclosure of contributions and lobbying relationships, with provisions to prevent the obfuscation of the funding source. Self-interested advocacy should be transparently reported whether or not it was AI-assisted. Meanwhile, we should increase penalties for organizations that benefit from AI-assisted impersonation of constituents in political processes, and set a greater expectation of responsibility to avoid “unknowing” use of these tools on their behalf.

Our most important recommendation is less legal and more cultural. Rather than trying to make it harder for AI to participate in the political process, make it easier for humans to do so.

The best way to fight an AI that can lobby for moneyed interests is to help the little guy lobby for theirs. Promote inclusion and engagement in the political process so that organic constituent communications grow alongside the potential growth of AI-directed communications. Encourage direct contact that generates more-than-digital relationships between constituents and their representatives, which will be an enduring way to privilege human stakeholders. Provide paid leave to allow people to vote as well as to testify before their legislature and participate in local town meetings and other civic functions. Provide childcare and accessible facilities at civic functions so that more community members can participate.

The threat of AI hacking our democracy is legitimate and concerning, but its solutions are consistent with our democratic values. Many of the ideas above are good governance reforms already being pushed and fought over at the federal and state level.

We don’t need to reinvent our democracy to save it from AI. We just need to continue the work of building a just and equitable political system. Hopefully ChatGPT will give us all some impetus to do that work faster.

This essay was written with Nathan Sanders, and appeared on the Belfer Center blog.

Posted on February 17, 2023 at 7:33 AM • View Comments

AI and Political Lobbying

Launched just weeks ago, ChatGPT is already threatening to upend how we draft everyday communications like emails, college essays and myriad other forms of writing.

Created by the company OpenAI, ChatGPT is a chatbot that can automatically respond to written prompts in a manner that is sometimes eerily close to human.

But for all the consternation over the potential for humans to be replaced by machines in formats like poetry and sitcom scripts, a far greater threat looms: artificial intelligence replacing humans in the democratic processes—not through voting, but through lobbying.

ChatGPT could automatically compose comments submitted in regulatory processes. It could write letters to the editor for publication in local newspapers. It could comment on news articles, blog entries and social media posts millions of times every day. It could mimic the work that the Russian Internet Research Agency did in its attempt to influence our 2016 elections, but without the agency’s reported multimillion-dollar budget and hundreds of employees.

Automatically generated comments aren’t a new problem. For some time, we have struggled with bots, machines that automatically post content. Five years ago, at least a million automatically drafted comments were believed to have been submitted to the Federal Communications Commission regarding proposed regulations on net neutrality. In 2019, a Harvard undergraduate, as a test, used a text-generation program to submit 1,001 comments in response to a government request for public input on a Medicaid issue. Back then, submitting comments was just a game of overwhelming numbers.

Platforms have gotten better at removing “coordinated inauthentic behavior.” Facebook, for example, has been removing over a billion fake accounts a year. But such messages are just the beginning. Rather than flooding legislators’ inboxes with supportive emails, or dominating the Capitol switchboard with synthetic voice calls, an AI system with the sophistication of ChatGPT but trained on relevant data could selectively target key legislators and influencers to identify the weakest points in the policymaking system and ruthlessly exploit them through direct communication, public relations campaigns, horse trading or other points of leverage.

When we humans do these things, we call it lobbying. Successful agents in this sphere pair precision message writing with smart targeting strategies. Right now, the only thing stopping a ChatGPT-equipped lobbyist from executing something resembling a rhetorical drone warfare campaign is a lack of precision targeting. AI could provide techniques for that as well.

A system that can understand political networks, if paired with the textual-generation capabilities of ChatGPT, could identify the member of Congress with the most leverage over a particular policy area—say, corporate taxation or military spending. Like human lobbyists, such a system could target undecided representatives sitting on committees controlling the policy of interest and then focus resources on members of the majority party when a bill moves toward a floor vote.

Once individuals and strategies are identified, an AI chatbot like ChatGPT could craft written messages to be used in letters, comments—anywhere text is useful. Human lobbyists could also target those individuals directly. It’s the combination that’s important: Editorial and social media comments only get you so far, and knowing which legislators to target isn’t itself enough.

This ability to understand and target actors within a network would create a tool for AI hacking, exploiting vulnerabilities in social, economic and political systems with incredible speed and scope. Legislative systems would be a particular target, because the motive for attacking policymaking systems is so strong, because the data for training such systems is so widely available and because the use of AI may be so hard to detect—particularly if it is being used strategically to guide human actors.

The data necessary to train such strategic targeting systems will only grow with time. Open societies generally make their democratic processes a matter of public record, and most legislators are eager—at least, performatively so—to accept and respond to messages that appear to be from their constituents.

Maybe an AI system could uncover which members of Congress have significant sway over leadership but still have low enough public profiles that there is only modest competition for their attention. It could then pinpoint the SuperPAC or public interest group with the greatest impact on that legislator’s public positions. Perhaps it could even calibrate the size of donation needed to influence that organization or direct targeted online advertisements carrying a strategic message to its members. For each policy end, the right audience; and for each audience, the right message at the right time.

What makes the threat of AI-powered lobbyists greater than the threat already posed by the high-priced lobbying firms on K Street is their potential for acceleration. Human lobbyists rely on decades of experience to find strategic solutions to achieve a policy outcome. That expertise is limited, and therefore expensive.

AI could, theoretically, do the same thing much more quickly and cheaply. Speed out of the gate is a huge advantage in an ecosystem in which public opinion and media narratives can become entrenched quickly, as is being nimble enough to shift rapidly in response to chaotic world events.

Moreover, the flexibility of AI could help achieve influence across many policies and jurisdictions simultaneously. Imagine an AI-assisted lobbying firm that can attempt to place legislation in every single bill moving in the US Congress, or even across all state legislatures. Lobbying firms tend to work within one state only, because there are such complex variations in law, procedure and political structure. With AI assistance in navigating these variations, it may become easier to exert power across political boundaries.

Just as teachers will have to change how they give students exams and essay assignments in light of ChatGPT, governments will have to change how they relate to lobbyists.

To be sure, there may also be benefits to this technology in the democracy space; the biggest one is accessibility. Not everyone can afford an experienced lobbyist, but a software interface to an AI system could be made available to anyone. If we’re lucky, maybe this kind of strategy-generating AI could revitalize the democratization of democracy by giving this kind of lobbying power to the powerless.

However, the biggest and most powerful institutions will likely use any AI lobbying techniques most successfully. After all, executing the best lobbying strategy still requires insiders—people who can walk the halls of the legislature—and money. Lobbying isn’t just about giving the right message to the right person at the right time; it’s also about giving money to the right person at the right time. And while an AI chatbot can identify who should be on the receiving end of those campaign contributions, humans will, for the foreseeable future, need to supply the cash. So while it’s impossible to predict what a future filled with AI lobbyists will look like, it will probably make the already influential and powerful even more so.

This essay was written with Nathan Sanders, and previously appeared in the New York Times.

Edited to Add: After writing this, we discovered that a research group is researching AI and lobbying:

We used autoregressive large language models (LLMs, the same type of model behind the now wildly popular ChatGPT) to systematically conduct the following steps. (The full code is available at this GitHub link: https://github.com/JohnNay/llm-lobbyist.)

  1. Summarize official U.S. Congressional bill summaries that are too long to fit into the context window of the LLM so the LLM can conduct steps 2 and 3.
  2. Using either the original official bill summary (if it was not too long), or the summarized version:
    1. Assess whether the bill may be relevant to a company based on a company’s description in its SEC 10K filing.
    2. Provide an explanation for why the bill is relevant or not.
    3. Provide a confidence level to the overall answer.
  3. If the bill is deemed relevant to the company by the LLM, draft a letter to the sponsor of the bill arguing for changes to the proposed legislation.

Here is the paper.

EDITED TO ADD (9/12): Emily Bender has a critique of this essay.

Posted on January 18, 2023 at 7:19 AM • View Comments

Recreating Democracy

Last week, I hosted a two-day workshop on recreating democracy.

The idea was to bring together people from a variety of disciplines who are all thinking about different aspects of democracy, less from a “what we need to do today” perspective and more from a blue-sky future perspective. My remit to the participants was this:

The idea is to start from scratch, to pretend we’re forming a new country and don’t have any precedent to deal with. And that we don’t have any unique interests to perturb our thinking. The modern representative democracy was the best form of government mid-eighteenth century politicians technology could invent. The twenty-first century is a very different place technically, scientifically, and philosophically. What could democracy look like if it were reinvented today? Would it even be democracy­—what comes after democracy?

Some questions to think about:

  • Representative democracies were built under the assumption that travel and communications were difficult. Does it still make sense to organize our representative units by geography? Or to send representatives far away to create laws in our name? Is there a better way for people to choose collective representatives?
  • Indeed, the very idea of representative government is due to technological limitations. If an AI system could find the optimal solution for balancing every voter’s preferences, would it still make sense to have representatives­—or should we vote for ideas and goals instead?
  • With today’s technology, we can vote anywhere and any time. How should we organize the temporal pattern of voting—­and of other forms of participation?
  • Starting from scratch, what is today’s ideal government structure? Does it make sense to have a singular leader “in charge” of everything? How should we constrain power­—is there something better than the legislative/judicial/executive set of checks and balances?
  • The size of contemporary political units ranges from a few people in a room to vast nation-states and alliances. Within one country, what might the smaller units be­—and how do they relate to one another?
  • Who has a voice in the government? What does “citizen” mean? What about children? Animals? Future people (and animals)? Corporations? The land?
  • And much more: What about the justice system? Is the twelfth-century jury form still relevant? How do we define fairness? Limit financial and military power? Keep our system robust to psychological manipulation?

My perspective, of course, is security. I want to create a system that is resilient against hacking: one that can evolve as both technologies and threats evolve.

The format was one that I have used before. Forty-eight people meet over two days. There are four ninety-minute panels per day, with six people on each. Everyone speaks for ten minutes, and the rest of the time is devoted to questions and comments. Ten minutes means that no one gets bogged down in jargon or details. Long breaks between sessions and evening dinners allow people to talk more informally. The result is a very dense, idea-rich environment that I find extremely valuable. (See the first ten comments below for details of the conversations.)

It was amazing event. Everyone participated. Everyone was interesting. (Details of the event—emerging themes, notes from the speakers—are in the comments.) It’s a week later and I am still buzzing with ideas. I hope this is only the first of an ongoing series of similar workshops.

Posted on December 14, 2022 at 9:30 PM • View Comments

Information Flows and Democracy

Henry Farrell and I published a paper on fixing American democracy: “Rechanneling Beliefs: How Information Flows Hinder or Help Democracy.”

It’s much easier for democratic stability to break down than most people realize, but this doesn’t mean we must despair over the future. It’s possible, though very difficult, to back away from our current situation towards one of greater democratic stability. This wouldn’t entail a restoration of a previous status quo. Instead, it would recognize that the status quo was less stable than it seemed, and a major source of the tensions that have started to unravel it. What we need is a dynamic stability, one that incorporates new forces into American democracy rather than trying to deny or quash them.

This paper is our attempt to explain what this might mean in practice. We start by analyzing the problem and explaining more precisely why a breakdown in public consensus harms democracy. We then look at how these beliefs are being undermined by three feedback loops, in which anti-democratic actions and anti-democratic beliefs feed on each other. Finally, we explain how these feedback loops might be redirected so as to sustain democracy rather than undermining it.

To be clear: redirecting these and other energies in more constructive ways presents enormous challenges, and any plausible success will at best be untidy and provisional. But, almost by definition, that’s true of any successful democratic reforms where people of different beliefs and values need to figure out how to coexist. Even when it’s working well, democracy is messy. Solutions to democratic breakdowns are going to be messy as well.

This is part of our series of papers looking at democracy as an information system. The first paper was “Common-Knowledge Attacks on Democracy.”

Posted on June 9, 2021 at 6:46 AM • View Comments

Should There Be Limits on Persuasive Technologies?

Persuasion is as old as our species. Both democracy and the market economy depend on it. Politicians persuade citizens to vote for them, or to support different policy positions. Businesses persuade consumers to buy their products or services. We all persuade our friends to accept our choice of restaurant, movie, and so on. It’s essential to society; we couldn’t get large groups of people to work together without it. But as with many things, technology is fundamentally changing the nature of persuasion. And society needs to adapt its rules of persuasion or suffer the consequences.

Democratic societies, in particular, are in dire need of a frank conversation about the role persuasion plays in them and how technologies are enabling powerful interests to target audiences. In a society where public opinion is a ruling force, there is always a risk of it being mobilized for ill purposes—­such as provoking fear to encourage one group to hate another in a bid to win office, or targeting personal vulnerabilities to push products that might not benefit the consumer.

In this regard, the United States, already extremely polarized, sits on a precipice.

There have long been rules around persuasion. The US Federal Trade Commission enforces laws that claims about products “must be truthful, not misleading, and, when appropriate, backed by scientific evidence.” Political advertisers must identify themselves in television ads. If someone abuses a position of power to force another person into a contract, undue influence can be argued to nullify that agreement. Yet there is more to persuasion than the truth, transparency, or simply applying pressure.

Persuasion also involves psychology, and that has been far harder to regulate. Using psychology to persuade people is not new. Edward Bernays, a pioneer of public relations and nephew to Sigmund Freud, made a marketing practice of appealing to the ego. His approach was to tie consumption to a person’s sense of self. In his 1928 book Propaganda, Bernays advocated engineering events to persuade target audiences as desired. In one famous stunt, he hired women to smoke cigarettes while taking part in the 1929 New York City Easter Sunday parade, causing a scandal while linking smoking with the emancipation of women. The tobacco industry would continue to market lifestyle in selling cigarettes into the 1960s.

Emotional appeals have likewise long been a facet of political campaigns. In the 1860 US presidential election, Southern politicians and newspaper editors spread fears of what a “Black Republican” win would mean, painting horrific pictures of what the emancipation of slaves would do to the country. In the 2020 US presidential election, modern-day Republicans used Cuban Americans’ fears of socialism in ads on Spanish-language radio and messaging on social media. Because of the emotions involved, many voters believed the campaigns enough to let them influence their decisions.

The Internet has enabled new technologies of persuasion to go even further. Those seeking to influence others can collect and use data about targeted audiences to create personalized messaging. Tracking the websites a person visits, the searches they make online, and what they engage with on social media, persuasion technologies enable those who have access to such tools to better understand audiences and deliver more tailored messaging where audiences are likely to see it most. This information can be combined with data about other activities, such as offline shopping habits, the places a person visits, and the insurance they buy, to create a profile of them that can be used to develop persuasive messaging that is aimed at provoking a specific response.

Our senses of self, meanwhile, are increasingly shaped by our interaction with technology. The same digital environment where we read, search, and converse with our intimates enables marketers to take that data and turn it back on us. A modern day Bernays no longer needs to ferret out the social causes that might inspire you or entice you­—you’ve likely already shared that by your online behavior.

Some marketers posit that women feel less attractive on Mondays, particularly first thing in the morning—­and therefore that’s the best time to advertise cosmetics to them. The New York Times once experimented by predicting the moods of readers based on article content to better target ads, enabling marketers to find audiences when they were sad or fearful. Some music streaming platforms encourage users to disclose their current moods, which helps advertisers target subscribers based on their emotional states.

The phones in our pockets provide marketers with our location in real time, helping deliver geographically relevant ads, such as propaganda to those attending a political rally. This always-on digital experience enables marketers to know what we are doing­—and when, where, and how we might be feeling at that moment.

All of this is not intended to be alarmist. It is important not to overstate the effectiveness of persuasive technologies. But while many of them are more smoke and mirrors than reality, it is likely that they will only improve over time. The technology already exists to help predict moods of some target audiences, pinpoint their location at any given time, and deliver fairly tailored and timely messaging. How far does that ability need to go before it erodes the autonomy of those targeted to make decisions of their own free will?

Right now, there are few legal or even moral limits on persuasion­—and few answers regarding the effectiveness of such technologies. Before it is too late, the world needs to consider what is acceptable and what is over the line.

For example, it’s been long known that people are more receptive to advertisements made with people who look like them: in race, ethnicity, age, gender. Ads have long been modified to suit the general demographic of the television show or magazine they appear in. But we can take this further. The technology exists to take your likeness and morph it with a face that is demographically similar to you. The result is a face that looks like you, but that you don’t recognize. If that turns out to be more persuasive than coarse demographic targeting, is that okay?

Another example: Instead of just advertising to you when they detect that you are vulnerable, what if advertisers craft advertisements that deliberately manipulate your mood? In some ways, being able to place ads alongside content that is likely to provoke a certain emotional response enables advertisers to do this already. The only difference is that the media outlet claims it isn’t crafting the content to deliberately achieve this. But is it acceptable to actively prime a target audience and then to deliver persuasive messaging that fits the mood?

Further, emotion-based decision-making is not the rational type of slow thinking that ought to inform important civic choices such as voting. In fact, emotional thinking threatens to undermine the very legitimacy of the system, as voters are essentially provoked to move in whatever direction someone with power and money wants. Given the pervasiveness of digital technologies, and the often instant, reactive responses people have to them, how much emotion ought to be allowed in persuasive technologies? Is there a line that shouldn’t be crossed?

Finally, for most people today, exposure to information and technology is pervasive. The average US adult spends more than eleven hours a day interacting with media. Such levels of engagement lead to huge amounts of personal data generated and aggregated about you­—your preferences, interests, and state of mind. The more those who control persuasive technologies know about us, what we are doing, how we are feeling, when we feel it, and where we are, the better they can tailor messaging that provokes us into action. The unsuspecting target is grossly disadvantaged. Is it acceptable for the same services to both mediate our digital experience and to target us? Is there ever such thing as too much targeting?

The power dynamics of persuasive technologies are changing. Access to tools and technologies of persuasion is not egalitarian. Many require large amounts of both personal data and computation power, turning modern persuasion into an arms race where the better resourced will be better placed to influence audiences.

At the same time, the average person has very little information about how these persuasion technologies work, and is thus unlikely to understand how their beliefs and opinions might be manipulated by them. What’s more, there are few rules in place to protect people from abuse of persuasion technologies, much less even a clear articulation of what constitutes a level of manipulation so great it effectively takes agency away from those targeted. This creates a positive feedback loop that is dangerous for society.

In the 1970s, there was widespread fear about so-called subliminal messaging, which claimed that images of sex and death were hidden in the details of print advertisements, as in the curls of smoke in cigarette ads and the ice cubes of liquor ads. It was pretty much all a hoax, but that didn’t stop the Federal Trade Commission and the Federal Communications Commission from declaring it an illegal persuasive technology. That’s how worried people were about being manipulated without their knowledge and consent.

It is time to have a serious conversation about limiting the technologies of persuasion. This must begin by articulating what is permitted and what is not. If we don’t, the powerful persuaders will become even more powerful.

This essay was written with Alicia Wanless, and previously appeared in Foreign Policy.

EDITED TO ADD: Ukrainian translation.

Posted on December 14, 2020 at 2:03 PM • View Comments

Undermining Democracy

Last Thursday, Rudy Giuliani, a Trump campaign lawyer, alleged a widespread voting conspiracy involving Venezuela, Cuba, and China. Another lawyer, Sidney Powell, argued that Mr. Trump won in a landslide, the entire election in swing states should be overturned and the legislatures should make sure that the electors are selected for the president.

The Republican National Committee swung in to support her false claim that Mr. Trump won in a landslide, while Michigan election officials have tried to stop the certification of the vote.

It is wildly unlikely that their efforts can block Joe Biden from becoming president. But they may still do lasting damage to American democracy for a shocking reason: the moves have come from trusted insiders.

American democracy’s vulnerability to disinformation has been very much in the news since the Russian disinformation campaign in 2016. The fear is that outsiders, whether they be foreign or domestic actors, will undermine our system by swaying popular opinion and election results.

This is half right. American democracy is an information system, in which the information isn’t bits and bytes but citizens’ beliefs. When peoples’ faith in the democratic system is undermined, democracy stops working. But as information security specialists know, outsider attacks are hard. Russian trolls, who don’t really understand how American politics works, have actually had a difficult time subverting it.

When you really need to worry is when insiders go bad. And that is precisely what is happening in the wake of the 2020 presidential election. In traditional information systems, the insiders are the people who have both detailed knowledge and high level access, allowing them to bypass security measures and more effectively subvert systems. In democracy, the insiders aren’t just the officials who manage voting but also the politicians who shape what people believe about politics. For four years, Donald Trump has been trying to dismantle our shared beliefs about democracy. And now, his fellow Republicans are helping him.

Democracy works when we all expect that votes will be fairly counted, and defeated candidates leave office. As the democratic theorist Adam Przeworski puts it, democracy is “a system in which parties lose elections.” These beliefs can break down when political insiders make bogus claims about general fraud, trying to cling to power when the election has gone against them.

It’s obvious how these kinds of claims damage Republican voters’ commitment to democracy. They will think that elections are rigged by the other side and will not accept the judgment of voters when it goes against their preferred candidate. Their belief that the Biden administration is illegitimate will justify all sorts of measures to prevent it from functioning.

It’s less obvious that these strategies affect Democratic voters’ faith in democracy, too. Democrats are paying attention to Republicans’ efforts to stop the votes of Democratic voters ­- and especially Black Democratic voters -­ from being counted. They, too, are likely to have less trust in elections going forward, and with good reason. They will expect that Republicans will try to rig the system against them. Mr. Trump is having a hard time winning unfairly, because he has lost in several states. But what if Mr. Biden’s margin of victory depended only on one state? What if something like that happens in the next election?

The real fear is that this will lead to a spiral of distrust and destruction. Republicans ­ who are increasingly committed to the notion that the Democrats are committing pervasive fraud -­ will do everything that they can to win power and to cling to power when they can get it. Democrats ­- seeing what Republicans are doing ­ will try to entrench themselves in turn. They suspect that if the Republicans really win power, they will not ever give it back. The claims of Republicans like Senator Mike Lee of Utah that America is not really a democracy might become a self-fulfilling prophecy.

More likely, this spiral will not directly lead to the death of American democracy. The U.S. federal system of government is complex and hard for any one actor or coalition to dominate completely. But it may turn American democracy into an unworkable confrontation between two hostile camps, each unwilling to make any concession to its adversary.

We know how to make voting itself more open and more secure; the literature is filled with vital and important suggestions. The more difficult problem is this. How do you shift the collective belief among Republicans that elections are rigged?

Political science suggests that partisans are more likely to be persuaded by fellow partisans, like Brad Raffensperger, the Republican secretary of state in Georgia, who said that election fraud wasn’t a big problem. But this would only be effective if other well-known Republicans supported him.

Public outrage, alternatively, can sometimes force officials to back down, as when people crowded in to denounce the Michigan Republican election officials who were trying to deny certification of their votes.

The fundamental problem, however, is Republican insiders who have convinced themselves that to keep and hold power, they need to trash the shared beliefs that hold American democracy together.

They may have long-term worries about the consequences, but they’re unlikely to do anything about those worries in the near-term unless voters, wealthy donors or others whom they depend on make them pay short-term costs.

This essay was written with Henry Farrell, and previously appeared in the New York Times.

Posted on November 27, 2020 at 6:10 AM • View Comments

Artificial Personas and Public Discourse

Presidential campaign season is officially, officially, upon us now, which means it’s time to confront the weird and insidious ways in which technology is warping politics. One of the biggest threats on the horizon: artificial personas are coming, and they’re poised to take over political debate. The risk arises from two separate threads coming together: artificial intelligence-driven text generation and social media chatbots. These computer-generated “people” will drown out actual human discussions on the Internet.

Text-generation software is already good enough to fool most people most of the time. It’s writing news stories, particularly in sports and finance. It’s talking with customers on merchant websites. It’s writing convincing op-eds on topics in the news (though there are limitations). And it’s being used to bulk up “pink-slime journalism”—websites meant to appear like legitimate local news outlets but that publish propaganda instead.

There’s a record of algorithmic content pretending to be from individuals, as well. In 2017, the Federal Communications Commission had an online public-commenting period for its plans to repeal net neutrality. A staggering 22 million comments were received. Many of them—maybe half—were fake, using stolen identities. These comments were also crude; 1.3 million were generated from the same template, with some words altered to make them appear unique. They didn’t stand up to even cursory scrutiny.

These efforts will only get more sophisticated. In a recent experiment, Harvard senior Max Weiss used a text-generation program to create 1,000 comments in response to a government call on a Medicaid issue. These comments were all unique, and sounded like real people advocating for a specific policy position. They fooled the Medicaid.gov administrators, who accepted them as genuine concerns from actual human beings. This being research, Weiss subsequently identified the comments and asked for them to be removed, so that no actual policy debate would be unfairly biased. The next group to try this won’t be so honorable.

Chatbots have been skewing social-media discussions for years. About a fifth of all tweets about the 2016 presidential election were published by bots, according to one estimate, as were about a third of all tweets about that year’s Brexit vote. An Oxford Internet Institute report from last year found evidence of bots being used to spread propaganda in 50 countries. These tended to be simple programs mindlessly repeating slogans: a quarter million pro-Saudi “We all have trust in Mohammed bin Salman” tweets following the 2018 murder of Jamal Khashoggi, for example. Detecting many bots with a few followers each is harder than detecting a few bots with lots of followers. And measuring the effectiveness of these bots is difficult. The best analyses indicate that they did not affect the 2016 US presidential election. More likely, they distort people’s sense of public sentiment and their faith in reasoned political debate. We are all in the middle of a novel social experiment.

Over the years, algorithmic bots have evolved to have personas. They have fake names, fake bios, and fake photos—sometimes generated by AI. Instead of endlessly spewing propaganda, they post only occasionally. Researchers can detect that these are bots and not people, based on their patterns of posting, but the bot technology is getting better all the time, outpacing tracking attempts. Future groups won’t be so easily identified. They’ll embed themselves in human social groups better. Their propaganda will be subtle, and interwoven in tweets about topics relevant to those social groups.

Combine these two trends and you have the recipe for nonhuman chatter to overwhelm actual political speech.

Soon, AI-driven personas will be able to write personalized letters to newspapers and elected officials, submit individual comments to public rule-making processes, and intelligently debate political issues on social media. They will be able to comment on social-media posts, news sites, and elsewhere, creating persistent personas that seem real even to someone scrutinizing them. They will be able to pose as individuals on social media and send personalized texts. They will be replicated in the millions and engage on the issues around the clock, sending billions of messages, long and short. Putting all this together, they’ll be able to drown out any actual debate on the Internet. Not just on social media, but everywhere there’s commentary.

Maybe these persona bots will be controlled by foreign actors. Maybe it’ll be domestic political groups. Maybe it’ll be the candidates themselves. Most likely, it’ll be everybody. The most important lesson from the 2016 election about misinformation isn’t that misinformation occurred; it is how cheap and easy misinforming people was. Future technological improvements will make it all even more affordable.

Our future will consist of boisterous political debate, mostly bots arguing with other bots. This is not what we think of when we laud the marketplace of ideas, or any democratic political process. Democracy requires two things to function properly: information and agency. Artificial personas can starve people of both.

Solutions are hard to imagine. We can regulate the use of bots—a proposed California law would require bots to identify themselves—but that is effective only against legitimate influence campaigns, such as advertising. Surreptitious influence operations will be much harder to detect. The most obvious defense is to develop and standardize better authentication methods. If social networks verify that an actual person is behind each account, then they can better weed out fake personas. But fake accounts are already regularly created for real people without their knowledge or consent, and anonymous speech is essential for robust political debate, especially when speakers are from disadvantaged or marginalized communities. We don’t have an authentication system that both protects privacy and scales to the billions of users.

We can hope that our ability to identify artificial personas keeps up with our ability to disguise them. If the arms race between deep fakes and deep-fake detectors is any guide, that’ll be hard as well. The technologies of obfuscation always seem one step ahead of the technologies of detection. And artificial personas will be designed to act exactly like real people.

In the end, any solutions have to be nontechnical. We have to recognize the limitations of online political conversation, and again prioritize face-to-face interactions. These are harder to automate, and we know the people we’re talking with are actual people. This would be a cultural shift away from the internet and text, stepping back from social media and comment threads. Today that seems like a completely unrealistic solution.

Misinformation efforts are now common around the globe, conducted in more than 70 countries. This is the normal way to push propaganda in countries with authoritarian leanings, and it’s becoming the way to run a political campaign, for either a candidate or an issue.

Artificial personas are the future of propaganda. And while they may not be effective in tilting debate to one side or another, they easily drown out debate entirely. We don’t know the effect of that noise on democracy, only that it’ll be pernicious, and that it’s inevitable.

This essay previously appeared in TheAtlantic.com.

EDITED TO ADD: Jamie Susskind wrote a similar essay.

EDITED TO ADD (3/16): This essay has been translated into Spanish.

EDITED TO ADD (6/4): This essay has been translated into Portuguese.

Posted on January 13, 2020 at 8:21 AM • View Comments

Defending Democracies Against Information Attacks

To better understand influence attacks, we proposed an approach that models democracy itself as an information system and explains how democracies are vulnerable to certain forms of information attacks that autocracies naturally resist. Our model combines ideas from both international security and computer security, avoiding the limitations of both in explaining how influence attacks may damage democracy as a whole.

Our initial account is necessarily limited. Building a truly comprehensive understanding of democracy as an information system will be a Herculean labor, involving the collective endeavors of political scientists and theorists, computer scientists, scholars of complexity, and others.

In this short paper, we undertake a more modest task: providing policy advice to improve the resilience of democracy against these attacks. Specifically, we can show how policy makers not only need to think about how to strengthen systems against attacks, but also need to consider how these efforts intersect with public beliefs­—or common political knowledge­—about these systems, since public beliefs may themselves be an important vector for attacks.

In democracies, many important political decisions are taken by ordinary citizens (typically, in electoral democracies, by voting for political representatives). This means that citizens need to have some shared understandings about their political system, and that the society needs some means of generating shared information regarding who their citizens are and what they want. We call this common political knowledge, and it is largely generated through mechanisms of social aggregation (and the institutions that implement them), such as voting, censuses, and the like. These are imperfect mechanisms, but essential to the proper functioning of democracy. They are often compromised or non-existent in autocratic regimes, since they are potentially threatening to the rulers.

In modern democracies, the most important such mechanism is voting, which aggregates citizens’ choices over competing parties and politicians to determine who is to control executive power for a limited period. Another important mechanism is the census process, which play an important role in the US and in other democracies, in providing broad information about the population, in shaping the electoral system (through the allocation of seats in the House of Representatives), and in policy making (through the allocation of government spending and resources). Of lesser import are public commenting processes, through which individuals and interest groups can comment on significant public policy and regulatory decisions.

All of these systems are vulnerable to attack. Elections are vulnerable to a variety of illegal manipulations, including vote rigging. However, many kinds of manipulation are currently legal in the US, including many forms of gerrymandering, gimmicking voting time, allocating polling booths and resources so as to advantage or disadvantage particular populations, imposing onerous registration and identity requirements, and so on.

Censuses may be manipulated through the provision of bogus information or, more plausibly, through the skewing of policy or resources so that some populations are undercounted. Many of the political battles over the census over the past few decades have been waged over whether the census should undertake statistical measures to counter undersampling bias for populations who are statistically less likely to return census forms, such as minorities and undocumented immigrants. Current efforts to include a question about immigration status may make it less likely that undocumented or recent immigrants will return completed forms.

Finally, public commenting systems too are vulnerable to attacks intended to misrepresent the support for or opposition to specific proposals, including the formation of astroturf (artificial grassroots) groups and the misuse of fake or stolen identities in large-scale mail, fax, email or online commenting systems.

All these attacks are relatively well understood, even if policy choices might be improved by a better understanding of their relationship to shared political knowledge. For example, some voting ID requirements are rationalized through appeals to security concerns about voter fraud. While political scientists have suggested that these concerns are largely unwarranted, we currently lack a framework for evaluating the trade-offs, if any. Computer security concepts such as confidentiality, integrity, and availability could be combined with findings from political science and political theory to provide such a framework.

Even so, the relationship between social aggregation institutions and public beliefs is far less well understood by policy makers. Even when social aggregation mechanisms and institutions are robust against direct attacks, they may be vulnerable to more indirect attacks aimed at destabilizing public beliefs about them.

Democratic societies are vulnerable to (at least) two kinds of knowledge attacks that autocratic societies are not. First are flooding attacks that create confusion among citizens about what other citizens believe, making it far more difficult for them to organize among themselves. Second are confidence attacks. These attempt to undermine public confidence in the institutions of social aggregation, so that their results are no longer broadly accepted as legitimate representations of the citizenry.

Most obviously, democracies will function poorly when citizens do not believe that voting is fair. This makes democracies vulnerable to attacks aimed at destabilizing public confidence in voting institutions. For example, some of Russia’s hacking efforts against the 2016 presidential election were designed to undermine citizens’ confidence in the result. Russian hacking attacks against Ukraine, which targeted the systems through which election results were reported out, were intended to create confusion among voters about what the outcome actually was. Similarly, the “Guccifer 2.0” hacking identity, which has been attributed to Russian military intelligence, sought to suggest that the US electoral system had been compromised by the Democrats in the days immediately before the presidential vote. If, as expected, Donald Trump had lost the election, these claims could have been combined with the actual evidence of hacking to create the appearance that the election was fundamentally compromised.

Similar attacks against the perception of fairness are likely to be employed against the 2020 US census. Should efforts to include a citizenship question fail, some political actors who are disadvantaged by demographic changes such as increases in foreign-born residents and population shift from rural to urban and suburban areas will mount an effort to delegitimize the census results. Again, the genuine problems with the census, which include not only the citizenship question controversy but also serious underfunding, may help to bolster these efforts.

Mechanisms that allow interested actors and ordinary members of the public to comment on proposed policies are similarly vulnerable. For example, the Federal Communication Commission (FCC) announced in 2017 that it was proposing to repeal its net neutrality ruling. Interest groups backing the FCC rollback correctly anticipated a widespread backlash from a politically active coalition of net neutrality supporters. The result was warfare through public commenting. More than 22 million comments were filed, most of which appeared to be either automatically generated or form letters. Millions of these comments were apparently fake, and attached unsuspecting people’s names and email addresses to comments supporting the FCC’s repeal efforts. The vast majority of comments that were not either form letters or automatically generated opposed the FCC’s proposed ruling. The furor around the commenting process was magnified by claims from inside the FCC (later discredited) that the commenting process had also been subjected to a cyberattack.

We do not yet know the identity and motives of the actors behind the flood of fake comments, although the New York State Attorney-General’s office has issued subpoenas for records from a variety of lobbying and advocacy organizations. However, by demonstrating that the commenting process was readily manipulated, the attack made it less likely that the apparently genuine comments of those opposing the FCC’s proposed ruling would be treated as useful evidence of what the public believed. The furor over purported cyberattacks, and the FCC’s unwillingness itself to investigate the attack, have further undermined confidence in an online commenting system that was intended to make the FCC more open to the US public.

We do not know nearly enough about how democracies function as information systems. Generating a better understanding is itself a major policy challenge, which will require substantial resources and, even more importantly, common understandings and shared efforts across a variety of fields of knowledge that currently don’t really engage with each other.

However, even this basic sketch of democracy’s informational aspects can provide policy makers with some key lessons. The most important is that it may be as important to bolster shared public beliefs about key institutions such as voting, public commenting, and census taking against attack, as to bolster the mechanisms and related institutions themselves.

Specifically, many efforts to mitigate attacks against democratic systems begin with spreading public awareness and alarm about their vulnerabilities. This has the benefit of increasing awareness about real problems, but it may ­ especially if exaggerated for effect ­ damage public confidence in the very social aggregation institutions it means to protect. This may mean, for example, that public awareness efforts about Russian hacking that are based on flawed analytic techniques may themselves damage democracy by exaggerating the consequences of attacks.

More generally, this poses important challenges for policy efforts to secure social aggregation institutions against attacks. How can one best secure the systems themselves without damaging public confidence in them? At a minimum, successful policy measures will not simply identify problems in existing systems, but provide practicable, publicly visible, and readily understandable solutions to mitigate them.

We have focused on the problem of confidence attacks in this short essay, because they are both more poorly understood and more profound than flooding attacks. Given historical experience, democracy can probably survive some amount of disinformation about citizens’ beliefs better than it can survive attacks aimed at its core institutions of aggregation. Policy makers need a better understanding of the relationship between political institutions and social beliefs: specifically, the importance of the social aggregation institutions that allow democracies to understand themselves.

There are some low-hanging fruit. Very often, hardening these institutions against attacks on their confidence will go hand in hand with hardening them against attacks more generally. Thus, for example, reforms to voting that require permanent paper ballots and random auditing would not only better secure voting against manipulation, but would have moderately beneficial consequences for public beliefs too.

There are likely broadly similar solutions for public commenting systems. Here, the informational trade-offs are less profound than for voting, since there is no need to balance the requirement for anonymity (so that no-one can tell who voted for who ex post) against other requirements (to ensure that no-one votes twice or more, no votes are changed and so on). Instead, the balance to be struck is between general ease of access and security, making it easier, for example, to leverage secondary sources to validate identity.

Both the robustness of and public confidence in the US census and the other statistical systems that guide the allocation of resources could be improved by insulating them better from political control. For example, a similar system could be used to appoint the director of the census to that for the US Comptroller-General, requiring bipartisan agreement for appointment, and making it hard to exert post-appointment pressure on the official.

Our arguments also illustrate how some well-intentioned efforts to combat social influence operations may have perverse consequences for general social beliefs. The perception of security is at least as important as the reality of security, and any defenses against information attacks need to address both.

However, we need far better developed intellectual tools if we are to properly understand the trade-offs, instead of proposing clearly beneficial policies, and avoiding straightforward mistakes. Forging such tools will require computer security specialists to start thinking systematically about public beliefs as an integral part of the systems that they seek to defend. It will mean that more military oriented cybersecurity specialists need to think deeply about the functioning of democracy and the capacity of internal as well as external actors to disrupt it, rather than reaching for their standard toolkit of state-level deterrence tools. Finally, specialists in the workings of democracy have to learn how to think about democracy and its trade-offs in specifically informational terms.

This essay was written with Henry Farrell, and has previously appeared on Defusing Disinfo.

Posted on April 30, 2019 at 6:59 AM • View Comments

1 … 3 4 5 6

Sidebar photo of Bruce Schneier by Joe MacInnis.