Friday Squid Blogging: How Squid Skin Distorts Light
New research.
As usual, you can also use this squid post to talk about the security stories in the news that I haven’t covered.
Comments
not important • July 5, 2025 6:34 PM
https://www.yahoo.com/news/taiwan-brainprobe-platform-uses-ai-000703302.html
=As per a report on Taipei Times, researchers at Taipei Veterans General Hospital (TVGH)
have developed an artificial intelligence-powered platform capable of helping diagnose
schizophrenia.
The tools, a first-of-its-kind in the world, mark a major leap forward in the world of
psychiatric care, long hampered by subjectivity and guesswork.
Named BrainProbe, the tool analyzes magnetic resonance imaging (MRI) scans to detect
structural and functional changes in the brain associated with schizophrenia.
With an accuracy rate of 91.7 percent, it’s the world’s first AI tool of its kind and a
dramatic departure from traditional methods that rely heavily on clinical interviews and
behavioral observation.
“The field of psychiatry has long hoped to identify objective biological markers that can help quantify the symptoms [of mental illness],” TVGH Medical AI Development Center deputy director Albert Yang was quoted as saying by Taipei Times.
Schizophrenia is a complex mental disorder characterized by hallucinations, delusions, and cognitive disruptions.
With that challenge in mind, Yang’s team turned to artificial intelligence. Using more than a decade of brain scan data from over 1,500 individuals, including both healthy and those diagnosed with schizophrenia, the AI tool was trained to detect subtle, early-stage changes invisible to the human eye.
More than just a diagnostic tool, BrainProbe can also track how the brain changes over
time.
“We hope this platform can be applied across different ethnic groups to enable more
accurate research,” Yang added.=
That is move to real objectivity in mental health diagnostic. Many security related vectors.
lurker • July 6, 2025 12:16 AM
The UK govt has told the Alan Turing Institute “the UK’s AI capabilities was “critical” to national security and should be at the core of the Alan Turing Institute’s activities.”
‘https://www.bbc.com/news/articles/cy7nppe5gkgo
Clive Robinson • July 6, 2025 11:42 AM
@ lurker,
With regards the “Alan Turing Institute”(ATI), it’s very probably “money down the drain” already… and switching to a larger AI focus will almost certainly not produce any worthwhile benefits.
Let’s put it this way, if the stories I’ve been told have any real truth in them… Then a visit by Musk’s old DOGiE minions would probably improve things a lot.
The ATI have sort of been doing Current AI for most of the past decade,
“But can you remember anything they’ve done in that time?”
No me neither. But I’m not exactly surprised, it’s management at best gave been ineffective, and after a decade it’s still “slumming it” “cap in hand” in other organisations premises. Currently it’s got a back office or two hanging off the back of The British Library (that you might remember had a major ITCSec disaster not so long ago).
A little over a year ago a very large number of employees –a little under half– wrote what was a letter of serious complaint with legal implications. Put simply the leadership were accused of “non inclusive behaviour” that would have caused major issues if it got into a Tribunal or Court to do with employment issues. The Leadership response was to try to force “around a third of the employees into redundancy. And things carried on going down hill with what was very likely unlawful employment behaviours.
As for UK Technology Secretary Peter Kyle’s letter of last Thusday (3rd of this month) all I can say is what the BBC has “reported” is not realy what the letter was realy all about as reported in several other places including,
https://www.politico.eu/article/uk-government-ai-institute-prioritize-security-defense/
In short for the letter was a demand for “action this day” calling for a further serious root and branch overhaul of the ATI and it’s direction. In effect calling for a significant refocusing of it’s work direction away from “health” and on to “defense” with basically a decapitation strike on the leadership team. In the letter to the chair of the ATI, the Minister apparently wrote that,
“[I]t remains clear that further action is needed to ensure the ATI meets its full potential”.
Which is a diplomatic way of saying “fall on your sword” or “we will publicly terminate with prejudice”. Or another way of saying “no cushy jobs, knighthoods, MBEs, OBEs or other honours for you lot.”
But… Not the coincidence of the letter to the “Orange One in Office” demanding that Europe spends 5% of GDP on military. With the unstated implication that Europe should give that 5% to the US Defence Industry.
The UK is in an awkward position of it’s own political making… With Brexit more or less destroying sone of the world leading Aero & Space technology companies and forcing them to be sold to European or US entities.
The UK Government hope is if it can “get in on the bottom floor of AI” then it can “get the elevator”… Well personally I see current AI LLM, LRM, and ML systems really are a pile of “slop draining into the basement” drowning all who actually did not get the elevator…
Let’s be blunt the current AI LLM, LRM and ML systems are about as close to useless as you can get for actually doing anything “Frontline Defence” related. Defence systems need to have “Fully Autonomous High Capability” whilst also being low weight and power and occupying little volume. Otherwise “a man in a rig” will probably be more effective if you can keep the brain alive…
anon • July 6, 2025 3:14 PM
Site based age verification (sfw)
Announcements about Site-based AV give parents the illusion that their kids are protected — when in reality, nothing changes. Porn will still be everywhere.
It’s an absolute joke — a placebo solution pushed by imbeciles who clearly don’t give a shit about minors.
Device-level parental controls have existed for years, and can actually block a million sites. But politicians can’t take credit for them. So instead of empowering parents, lawmakers give them a headline and a false promise.
Watch these pricks go on TV or social media and pat themselves on the back for “making the internet safer” — while anyone can still run a basic Google search and instantly find billions of porn images. Bold-faced liars.
The one “good” thing about AV is that it’s a clear sign of political incompetence. If your lawmakers passed this kind of law, you can be sure they’re either corrupt, lying to you, stupid — or some combination of all three. At least it makes them easier to spot.
It might make some sense — even if we’d still disagree (because there are better solutions, see below) — to enforce AV on major mainstream platforms to prevent accidental exposure to porn. But the hypocrites in charge are systematically and deliberately sparing those platforms.
Clive Robinson • July 6, 2025 3:28 PM
@ Bruce, ALL,
As you probably realise by now I think that current AI Language Models and the ML that supposedly makes them is,
“A near total dud that has wasted billions”
And no doubt will burn billions more.
I’ve also mentioned I think it is the most horrendous stealer of privacy there is and whilst probably not the ultimate surveillance tool, it is probably the most invasive currently.
I’ve mentioned what I think is the business plan of Alphabet / Google, Meta, Microsoft, and others, and I’ve called it the “BE Plan” for,
“Bedazzle, Beguile, Bewitch, Befriend, and BETRAY”
After the basic steps they’ve all been following.
However there is another aspect to AI, that is billions has been thrown at it and few –maybe 2%– actually want to play with it and even less want to work with it.
So making money on Current AI is not realistically practical. In fact it is not making money and nor is it ever likely to wherever there is intelligence and choice available to individual users.
So there is a bemadening stage to add where it is forced on people even when they scream “NO”. We don’t yet have the crime of “mind rape” but it is something we should consider as a society whilst we still have the choice.
But Current AI is not free, there is always at least two fees that have to be paid in the “rent seeking” world,
1, The increased price to the consumer.
2, The selling of the consumers privacy.
I’m not the only person who thinks this “bemadening” process is being forced upon us for “share holder value”,
https://www.honest-broker.com/p/the-force-feeding-of-ai-on-an-unwilling
Interestingly there is also the word “begorging” for force feeding… It used to be used for the process of “fattening up for the slaughter” of amoungst other creatures dormice and geese. The latter to force a hard fat liver on the bird so that it made good Foie gras (French for fat liver).
Further I’m known as a person who does not do Social media, Email or messaging, and I don’t use “Cloud” or XaaS because I don’t have an Internet connection to my computers and never will have…
Worse for the “rent seekers” I purchased stand alone versions of commercial software or I developed my own software. And I find ways to make it work for me. Something where FOSS comes in rather usefully. So yes I run WordStar 4, and Office 97 amongst other commercial software. Yes I’m aware it’s all got “vulnerabilities” but to exploit them you need access…
However… It’s becoming clear with Web Browsers and similar that we are being forced into the maws of “rent seekers” something I think will spill over into all popular software including FOSS over the next few years.
People should use “the madness that is Win 11” to change their habits and select software that’s neither cloud dependant or destroys your privacy.
Clive Robinson • July 7, 2025 3:22 AM
@ ALL,
First Work against the wall come AI
There is a debate / discussion going on about who’s job is likely to fet significantly changed / obliterated by current LLM and ML systems.
I’ve already pointed out that “Language Models” lack not just awareness of their environment but the inability to gain it. And it’s not just through “lack of agency” (though it does not help).
Thus non repetitive manual trades or skills are not going to get hit except “on the admin side”.
I’ve noted that there are many administrative “make work” jobs and logically they would be almost trivial to replace… But being “make work” usually means “there is a political –small or big P– reason for the jobs existence”.
However there are exceptions, we’ve already seen this in the decline of low end “guard labour” jobs as CCTV and facial recognition takes over… However those that remain are for “the physical aspects” of deterrent or where static systems are not possible. It’s unlikely this will change for a while because of legal reasons. Disable a human guard and that ranges from assault through kidnapping to murder, all of which are “serious crimes against the person” so have significant criminal penalties. However spray paint a CCTV, break it, or steal it, and those are generally “minor crimes against property” currently. Authority uses “asymmetry of power” in “Might is Right” assumptions in the use of human guard labour. However the legislative process on technology is way behind on this “might is right” asymmetry (expect this to change fairly quickly and with very broad scope).
I don’t really expect “true creative” work to be replaced but “generic or in a style” work will get hit quite hard as it’s fairly easy for LLM and ML systems to be “formulaic + noise”. Thus “copy work” is going to get hit and that effects generic architecture / design / engineering.
The argument against this is usually along the lines of “learned regulations and legislation” being a major and important part of the work.
Unfortunately that’s a faux-argument. It does not take long to realise that “Professions” such as Law, Accountancy, Medicine, architecture, and regulated design are all “formulaic” and it’s something that back in the 1980’s the idea of “Expert Systems” drove the nail in the lid well and truly.
But worse for corporations, nearly all lawful senior management actions are “formulaic” and “heavily regulated”… Thus ideal to be replaced by LLM and ML systems.
So who’s work is safe?
As I’ve mentioned “original creative work” is not something current language based systems are going to replace any time soon if ever due to “view of environment” in a meaningful way…
But somebody has made similar observations and said it in a much more creative way,
“Thesis: interesting work (i.e., work worth doing) is less amenable to the use of AI than … non-interesting work.”
https://remark.ing/rob/rob/Thesis-interesting-work-ie
It’s an essential point that the AGI Hypers and similar trying to keep the bubble inflated are desperately trying to stop people realising.
So if you want to keep your job, you kind of need to make a major part,
1, Bespoke creative.
2, Uniquely creative.
3, Fit human perception.
4, As free from formulaic as reasonably possible.
And those points kind of mean “do what Language Model AI can not do”.
If you feel your software can be written by an LLM in the otherwise interesting “mid point” where all the creative stuff happens, then I would politely suggest you need to “up your game” or “move jobs whilst you still can”.
Genocide buff • July 7, 2025 2:43 PM
Does anybody have the link to the “Evidence Task” genocide documentation? Thought it was on a Protonmail but the link is not showing up on search, go figure. The ICC has it but it’s nice to keep a personal copy in case it goes down the memory hole.
not important • July 7, 2025 7:05 PM
https://www.yahoo.com/news/memristor-based-system-chinese-scientists-181509182.html
=In a bid to overcome shortcomings in scientific computing, Chinese scientists have
unveiled a new approach to sorting data that promises both higher speed and lower energy
consumption. The system combines memristors—electronic components that mimic the memory function of the human brain—with an advanced sorting algorithm to process large amounts of information more efficiently.
Researchers say this method could help overcome performance bottlenecks in not just
computing but also artificial intelligence (AI), and hardware design, where rapidly
organizing and analyzing vast datasets is essential. Beyond AI, potential applications
for this technology include smart traffic systems that analyze images in real time and
financial services that require quick risk assessments.
Traditional computing systems rely on the Von Neumann architecture, which separates data
storage and processing functions, typically using a central processing unit (CPU) to
handle calculations.
Unlike ordinary resistors, which simply reduce the flow of electricity in a circuit,
memristors have the unique ability to remember how much electrical charge has passed through them. This memory function allows memristors to adjust their resistance based on previous activity, enabling them to act as both storage and processing components.
By combining these functions, memristors could eliminate the need to transfer data
between separate memory and processing units, potentially leading to faster and more
energy-efficient computing systems.
Traditional hardware sorting relies on CPUs, GPUs, or specialised chips that compare
numbers step by step using sorting algorithms. Instead, the new method uses memristors to perform iterative search-based sorting, finding minimum or maximum values without directly comparing each pair, which in turn saves both time and energy.=
Clive Robinson • July 8, 2025 6:54 AM
@ Bruce, ALL,
The fall out of war can reach far
The news from South Korea will not be good for most Northern Hemisphere traditional arms manufacturers. Though long term readers here will have probably seen it coming.
Rep. Yu Yong-weon of the SK People Power Party,
“welcomed the decision, citing the increasing vulnerability of helicopters to low-cost missile and drone attacks — a vulnerability highlighted by recent developments on the battlefield in Ukraine.
Yu said,
“The reconsideration of Apache procurement is a positive step”
“We need to invest in drones and other cutting-edge systems.”
“Drones and smart systems are redefining modern battlefields”
“Rather than clinging to expensive legacy platforms, we must invest in capabilities that reflect the future of warfare.”
The South Koreas being on the end of a peninsula into the South China Seas potentially facing attacking forces from all directions, thus have one of the largest reserves of military equipment.
In short,
“Agility in, rhinos out”.
But the South Koreans are not alone in this viewpoint,
Analysts cite the heavy losses of Russian helicopters to shoulder-fired missiles in Ukraine as a cautionary signal. Meanwhile, the U.S. Army has canceled its next-generation attack helicopter project, with similar reviews underway in Japan and Australia.
It would appear the fall out of the Russian forces is crashing down around the globe.
I suspect this is also going to effect 6th Generation military aircraft as well. Drones will get larger and load wise considerably cheaper becoming almost disposable, acting as lift vehicles for glide / standoff weapons. Kind of like a cruise missile with multiple independent payloads. Whilst the control systems will “stand well off” and keep humans as far out of the fight as possible, maybe not as far as “Sin City” but probably outside the war zone.
We’ve seen this unlawful “go for civilians” traditional form of terrorism practiced as “power projection” for some years now and I suspect it is only going to get worse.
lurker • July 8, 2025 1:43 PM
The beat goes on …
The US State Department said it is investigating an imposter who used artificial intelligence to impersonate Secretary of State Marco Rubio and contact three foreign ministers.
The “unknown actor” is alleged to have used an artificially-generated voice of Rubio to contact officials via the Signal messaging app …
‘https://www.bbc.com/news/articles/crrqkyyjewno
Clive Robinson • July 8, 2025 9:43 PM
@ lurker, ALL,
With regards,
“The US State Department said it is investigating an imposter who used artificial intelligence to impersonate Secretary of State Marco Rubio and contact three foreign ministers.”
I smell a rat…
Lets put it this way
“Why make it public in the way it has been?”
There are two reasons that immediately spring to mind,
1, Rubio like many Trump appointees has done something stupid. Thus is now looking for a way out…
2, For some reason someone is looking to discredit Signal…
@ ALL,
But consider how do you verify someone is who they claim to be in Signal or any other communications system now we are in this,
“Brave New World of AI fakes.”
The usual first “refuge cry” for those who are not sufficiently technically adept would probably be,
“Use a cryptographic protocol”.
Only nearly all Crypto systems are as bust as “8 character passwords” in this respect. And the others are easily open to human failings or significant impracticalities.
The real issue is what ever you do you have three entities,
1, The person
2, The communications system
3, The verification system
And they mostly do not work in a way where you actually get “secure verification”.
To “verify the person” needs some form of bio-metrics, and as such they really are neither sufficiently reliable or stable with regards time and the environment…
To verify the communications channel superficially appears to be simple but a study of the history of Public-Key systems should give anyone pause to think just how easy PKCerts are to fail.
Similar logic applies to any supposedly “secure verification” system. Put simply as they get scaled up the security levels effectively gets scaled down for both human and technical reasons.
Clive Robinson • July 9, 2025 6:10 AM
@ Bruce, ALL
eSIM insecurities found
https://security-explorations.com/esim-security.html
“The hack proves that our research on Java Card from 2019 did matter. Oracle indicated the vulnerabilities we reported to the company in 2019 were rather irrelevant (the company referred to them as “security concerns”) / did not affect their production Java Card VM. These are now proved to be real bugs.”
So a half decade or so old vulnerability that “Oracle” unsurprisingly poo-pooed/down played[1].
The rest of the information is a little technical, kind of the equivalent of talking about “how to pick the lock” to a barn full of prized horses.
However the thing to note is,
“we see no base for extending MNO trust beyond its own network”
With MNO meaning the “Mobile Network Operator” or “service provider” in other parlance.
Or put another way
“Do not use the mobile network for any security related activities”
So no “side channel authentication” or “financial or other transactions” etc…
As for secure messaging apps, well some of them are built on the assumption the MNO gives a trustworthy foundation that it now obviously does not.
[1] Oracle has a long and significant history of having multiple security vulnerabilities, but from the very top claiming knowingly and falsely that they had the best security… Even going as far at one point to threaten people with DMCA or equivalent violations if they produced or went looking for vulnerabilities in Oracle code.
not important • July 9, 2025 7:00 PM
@Clive said “Whilst the control systems will “stand well off” and keep humans as far out of the fight as possible, maybe not as far as “Sin City” but probably outside the war zone.”
Yeah, the far humans are from target, they less restricted psychologically to apply lethal force – see ‘On Killing’ By Col. Grossman.
You may find this interesting:
A Teenager Built Archimedes’ Mythical Death Ray—and It May Actually Work
https://www.yahoo.com/news/teenager-built-archimedes-mythical-death-135500277.html
=While no archaeological evidence proves that the death ray—also known as a heat ray—was
ever used, that doesn’t necessarily mean it didn’t happen. Sener explained that the
ancient death ray was made up of a series of large concave mirrors, positioned to focus
the Sun’s rays onto a single focal point and potentially hitting a target up to 1,000
feet away.
Sener crafted a miniature version of the concept. He used a heat lamp and four concave
mirrors, each pointed to hit a mark on a piece of cardboard. Presented at the 2023
Matthews Hall Annual Science Fair, he found that as he added mirrors, he increased the
temperature at the target location.
He believes his experiments show that a larger-scale version—such as the one described in an account of the battle of Syracuse from 214 to 212 BC, which is alleged to have helped set approaching Roman ships on fire—were conceivable. Ancient efforts may have either used mirrors or highly polished shields.=
not important • July 9, 2025 7:07 PM
https://www.yahoo.com/news/contributor-human-brain-doesnt-learn-100000423.html
=The biggest threat isn’t that these metaphors confuse us about how AI works, but that they mislead us about our own brains. During past technological revolutions, scientists, as well as popular culture, tended to explore the idea that the human brain could be understood as analogous to one new machine after another: a clock, a switchboard, a computer. The latest erroneous metaphor is that our brains are like AI systems.
the early 20th century “black box” model from behaviorist psychology claimed only visible
behavior mattered.
As a result, mental healthcare often focused on managing symptoms rather than understanding their emotional or biological causes.
One of the most interesting discoveries in neuroscience in the past two decades is the
“default mode network,” a group of brain regions that becomes active when we are
daydreaming and not focused on a specific task. This network has also been found to play a role in reflecting on the past, imagining and thinking about ourselves and others.
Unlike human memory, which evolved to forget, update and reshape memories based on myriad factors,
AI memory can be designed to store information with much less distortion or
forgetting. A life in which people outsource memory to a system that remembers almost
everything isn’t an extension of the self; it breaks from the very mechanisms that make
us human.
This outsourcing may be tempting because this technology seems human to us, but
AI learns, understands and sees the world in fundamentally different ways, and doesn’t truly experience pain, love or curiosity like we do.
The consequences of this ongoing confusion could be disastrous — not because AI is inherently harmful,
but because instead of shaping it into a tool that complements our human minds, we will allow it to reshape us in its own image.=
Clive Robinson • July 10, 2025 3:45 PM
@ Bruce, ALL,
More of the Xmas gift that keeps on giving.
It’s AMD this time and four “go faster stripe” faults that open up critical information leakage via side channels,
https://www.theregister.com/2025/07/09/amd_tsa_side_channel/
“Akin to Meltdown and Spectre, the Transient Scheduler Attack (TSA) comprises four vulnerabilities that AMD said it discovered while looking into a Microsoft report about microarchitectural leaks.”
Whilst AMD rate them as “medium to low severity”…
Trend Micro and CrowdStrike assess them as “critical”.
The difference being the view over exploitation requirements and data leaked,
“In AMD’s view, the TSAs affecting its chips are not exploitable via malicious websites, and would need to be executed many times in order to reliably exfiltrate any data.”
That said Microsoft indicate that so far there are no exploits. For how long this will remain true, and how hard the exploiting process will be both remain “unknown”…
I guess a pertinent question would be,
“When are these “go faster stripe” exploits ever going to stop?”
As it appears the CPU designers do not want to give up the “specmanship” speed advantages I guess the answer is simply,
“Never.”
Clive Robinson • July 10, 2025 7:09 PM
@ Bruce, ALL,
At last a real ROI on Current AI
This has realy put a grin on my face,
https://www.theregister.com/2025/07/10/ai_agents_automatically_steal_cryptocurrency/
Two Universities nearly poles apart have collaborated on an AI project that goes through WEB 3.0 Smart Contracts to work out how to in effect defraud them,
“Researchers with University College London (UCL) and the University of Sydney (USYD) in Australia have devised an AI agent that can autonomously discover and exploit vulnerabilities in so-called smart contracts.”
“Using AI models to generate exploits for cryptocurrency contract flaws appears to be a promising business model, though not necessarily a legal one.”
Begs the old statement those Crypto Bros used to regurgitate on faith,
“Code is law”
Is far from true, it’s just taking time for legislation to play “catch up”.
Worse with the quoted “Crime and Punishment” figures lookin rather more than a little sparse,
“The risk of imprisonment may shift calculations somewhat. But given the current regulatory climate in the US and an estimated cybercrime enforcement rate of 0.05 percent, it would be a small risk adjustment”
Suggests that the odds of getting away with it are good…
Clive Robinson • July 10, 2025 7:10 PM
@ Bruce, ALL,
At last a real ROI on Current AI
This has realy put a grin on my face,
https://www.theregister.com/2025/07/10/ai_agents_automatically_steal_cryptocurrency/
Two Universities nearly poles apart have collaborated on an AI project that goes through WEB 3.0 Smart Contracts to work out how to in effect defraud them,
“Researchers with University College London (UCL) and the University of Sydney (USYD) in Australia have devised an AI agent that can autonomously discover and exploit vulnerabilities in so-called smart contracts.”
“Using AI models to generate exploits for cryptocurrency contract flaws appears to be a promising business model, though not necessarily a legal one.”
Begs the old statement those Crypto Bros used to regurgitate on faith,
“Code is law”
Is far from true, it’s just taking time for legislation to play “catch up”.
Worse with the quoted “Crime and Punishment” figures lookin rather more than a little sparse,
“The risk of imprisonment may shift calculations somewhat. But given the current regulatory climate in the US and an estimated cybercrime enforcement rate of 0.05 percent, it would be a small risk adjustment”
Suggests that the odds of getting away with it, are fairly good…
Clive Robinson • July 11, 2025 2:52 AM
@ Bruce, ALL,
Productivity drops with use of AI in development
https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/
To be honest the results of,
“Core Result
When developers are allowed to use AI tools, they take 19% longer to complete issues—a significant slowdown that goes against developer beliefs and expert forecasts. This gap between perception and reality is striking: developers expected AI to speed them up by 24%, and even after experiencing the slowdown, they still believed AI had sped them up by 20%.”
Do not in the least surprise me as I’ve seen similar before, and will almost certainly see it again in the future as with all “toy technology introduction”.
Thus it’s something that might at some point show advantages, but not before significant changes are made.
fib • July 15, 2025 10:58 AM
Data breach that put up to 100,000 lives at risk revealed.
What it did not inform (or isn’t made public)
The name or rank of the person who leaked the data
Whether that person has faced disciplinary action, court-martial, or criminal charges
Whether the MoD considers it individual error, system failure, or both
The article never mentions any punishment or accountability for the person who caused the breach.
‘https://www.independent.co.uk/news/uk/home-news/mod-data-breach-superinjunction-afghanistan-taliban-b2789218.html
Clive Robinson • July 15, 2025 12:27 PM
@ fib,
With regards,
“The article never mentions any punishment or accountability for the person who caused the breach.”
From other sources I gather the answer is none or a little bit of retraining, and it is a chain of events not an individual event.
Put together from snippets,
A person was given what appeared to them to be a short list of names in spreadsheet form that “had to be verified”. So they sent out a copy to an appropriate checking agent as part of their assigned work.
Only the list was not what it appeared to be. That was it was a limited view into a much larger background list… Which ment in reality it was near a “full list” containing thousands of names that you could see by just changing the view…
The list then got sent to others etc and a member of the press got to hear about it or a copy of it and started their own “verification” process by “asking questions”.
At which point the panic started.
Due to the “nuclear option” the “superinjunction” is, people started removing information from what had been available to view.
As I’ve mentioned before there is,
1, Data / Information
2, Meta data / information about information
3, Meta meta data / information about meta data and data.
It’s the latter you need to get your head around. Because “a hole in data” or a “change in sequence”, time, pace, length etc etc can be just as informative, sometimes more so than the actual “information”.
There is a story about the First Gulf War, that the first people to know it was happening outside the “oh so secret” areas of US Government / Mil were the Pizza Delivery people that had more orders than they had ever had on a single night and they were all going to just a few “locations” that were “known” to be Mil / Gov places.
Well the sudden speedy deliveries out of the Pizza places was “Meta-Meta-Data” to anyone else trying to get a pizza that night.
So would have been the Credit Card info at the bank checking and payment center. Likewise if it was collected back then “phone call data”.
There is only so much you can do to keep such things hidden from public view or even confidential.
Meta-Meta-Data is beginning to scare people at the tops of certain agencies etc. Because it’s a,
“Damned if you do Damned if you don’t”
Type intelligence event.
To see why consider all aircraft not on “millitary activities” are supposed to have ADS-B transponders powered up at all times to prevent collisions etc.
Now if you are on “millitary activities” then yes your ADS-B can be turned off BUT Mil-ATC via NOTAMs and Civ-ATC are supposed to “keep other aircraft away” again to prevent collisions etc.
The problem is the ADS-B on civil aircraft paint a picture in the sky and on the displays of computers and web pages. So… If you divert Civil flights then it easily shows up as a “hole in ordinary activity”…
So anyone knowing “what’s on the ground” etc in those areas can make a “situational assessment” and in effect do “2+2=hmm that’s odd” or worse.
Meta-Meta-Data is one of the reasons that randomising data via encryption fails as it falls to “traffic analysis”, likewise randomising meta-data fails as it falls to Meta-Meta-Data analysis as well. Trying to randomise Meta-Meta-Data is let’s just say difficult if not impossible for a great many reasons, not least of which are the laws of nature (work requires energy differential and movement, work is inefficient, thus waste heat occurs, that can only end up dissipated in the environment at some point due to entropy otherwise work ceases as the differential diminishes).
lurker • July 17, 2025 12:38 AM
@fib, Clve
This case demonstrates head-in-the-sand attitude at high levels:
the govt obtained a super-injunction to “hide” the event -after- some of the data had already appeared on Facebook.
Clive Robinson • July 17, 2025 6:41 AM
@ lurker, fib,
With regards,
<
blockquote>”the govt obtained a super-injunction to “hide” the event -after- some of the data had already appeared on Facebook.”
<
blockquote>
The history of stopping “secrets” getting into the news has a long history in the UK…
First up there were various misuses of other legislation on the “stretch to fit you up” basis.
Then came the “Defence Of the Realm Act”(DORA) that gave us “D Notices” that whilst “advisory” often got treated by ministers and editors as some sort of pact. These kind of ended when certain journal editors just ignored “the invitations” and published. Not much later editors passed the story to a subsiduary entirely outside the UK and sometimes Commonwealth. So D Notices got kind of relegated.
There are various bits in both versions of the “Official Secrets Acts” but they became incredibly embarrassing in the “Mad Maggie” era due to her misuse to try to get prosecutions of things that were quite lawful and she cared not who she attempted to destroy (maybe a certain “King burger muncher” took lessons via the Ray-Gun 😉
But in more recent times with “footballer’s Wives” image issues the rise of civil injunctions has arisen…
Yes they are “civil law” but they have “contempt of court” hanging over them that is worse than “criminal law” when it comes to defending yourself (put simply there is no defence that a judge can not dismiss to find you guilty, the only restraint is their bringing the profession into disrepute, which as you can imagine suffers from “The Old Boy Network” issue.
But Super Injunctions are different, because they are “magical” and cast a pawl across the world,
The first bit of magic, is that unlike for ordinary injunctions the very fact that the process has been started is secret and must not be reported. This causes an issue that you can be unintentionally in contempt. Because if you have no knowledge of the procedings you can not know they are in progress and protected. This has a “chilling effect” on “free speech” and “honest reporting”.
Consider someone gets to hear that you might be investigating a story. They get a third party “to hint” that a super injunction is in play even when one is not.
The second bit of magic is an injunction is against certain things and each can be challenged and struck down. The super injunction due to “maintaining secrecy” is very broad as it’s the process that is protected. So even challenging a part does not get you anywhere you have to “destroy the whole” there is no “death by a thousand cuts” where you pick of a bit at a time.
Supposadly the requirments for a super injunction are much higher than for an ordinary injunction to stop the process from being abused… However “the crown and it’s agents” are automatically considered not to be abusers.
But also they have the “National Security” card they can play, in that they are not required to disclose facts to the court only approved court officers in closed session as “in camera”.
It is derived from a Latin phrase meaning “in private” or “in chambers” and it is seldom used in most legal cases because it is to call for proceedings conducted without public access. In the case of National Security it means that even a defendant and their legal representatives are refused access and sometimes even the judge… Who in effect gets directed by a “crown approved agent”. It is without doubt an abuse of basic principles of justice.
So given the low bar on entry given to the crown and the high bar on challenge to all others superinjunctions for “Official Secrecy” has a certain attraction. Add to that you are guilty no matter where you are in the world it kills off the likes of “Spy-Catcher” events where publishing was outside of legal jurisdiction thus could not be stopped and the usual “Blackmail by officialdom” was not going to work.
The fact that some of the information was already in the public domain was irrelevant…
As a third bit of magic is super injunctions effectively become “retrospective”… That is they protect the whole process in secrecy thus anything put up on say a blog prior to a super injunction being started or granted is covered and thus must be made “unpublic” in some way.
This obviously falls into the natural hole that Meta-Meta-Data is and it’s something that the legal profession has trouble getting it’s head around.
There is the old saw of,
“Nature abhorres a vacuum”
Well it’s even worse for Information systems. It’s now generally accepted in ICT Forensics that information is like a geological structure. You just can not remove or replace bits of information without creating sink holes and surface subsidence which almost always shows up rather embarrassingly.
Clive Robinson • July 18, 2025 3:08 PM
@ lurker, fib,
Further to my above…
A UK barrister released a video on YouTube subsequent to my posting,
https://m.youtube.com/watch?v=nhQTT4HsvDI
But also I’ve only just become aware of and seen. You might find it interesting.
It reveals a couple of things I was not aware of such as the Gov applied for an ordinary injunction, and it was the Court it’s self that upped it to a Super injunction.
Also the latin phrase for “against the world” scoping (no don’t try pronouncing it unless you are sure your teeth are firmly in place 😉
But the real eye opener for many is the £7billion cost or £350,000/person cost of “getting the effected people out of the trouble zone”.
However it should be noted that the bulk of the spending was not actually on the individuals directly… But in the manner of all bureaucracies on,
“The process and man power to facilitate it”.
Hope that helps get your head around the issues that have so far become public.
Subscribe to comments on this entry
Leave a comment
Sidebar photo of Bruce Schneier by Joe MacInnis.
XKCD • July 5, 2025 3:27 PM
Obligatory, but on point xckd.
https://xkcd.com/3109/