Security Risks of Government Hacking
Some of us — myself included — have proposed lawful government hacking as an alternative to backdoors. A new report from the Center of Internet and Society looks at the security risks of allowing government hacking. They include:
- Disincentive for vulnerability disclosure
- Cultivation of a market for surveillance tools
- Attackers co-opt hacking tools over which governments have lost control
- Attackers learn of vulnerabilities through government use of malware
- Government incentives to push for less-secure software and standards
- Government malware affects innocent users.
These risks are real, but I think they’re much less than mandating backdoors for everyone. From the report’s conclusion:
Government hacking is often lauded as a solution to the “going dark” problem. It is too dangerous to mandate encryption backdoors, but targeted hacking of endpoints could ensure investigators access to same or similar necessary data with less risk. Vulnerabilities will never affect everyone, contingent as they are on software, network configuration, and patch management. Backdoors, however, mean everybody is vulnerable and a security failure fails catastrophically. In addition, backdoors are often secret, while eventually, vulnerabilities will typically be disclosed and patched.
The key to minimizing the risks is to ensure that law enforcement (or whoever) report all vulnerabilities discovered through the normal process, and use them for lawful hacking during the period between reporting and patching. Yes, that’s a big ask, but the alternatives are worse.
This is the canonical lawful hacking paper.