$ 44.53 € 51.73 zł 11.95
+20° Kyiv +18° Warsaw +30° Washington

AI could make it harder for governments to use hacking tools — TechCrunch

Lev Shevtsov 31 August 2026 18:26
AI could make it harder for governments to use hacking tools — TechCrunch

The development of artificial intelligence could potentially reduce the number of software vulnerabilities and make it harder for law enforcement and intelligence agencies to hack targets for surveillance. Cryptography professor Matthew Green described this prospect earlier in August, TechCrunch reports.

According to Green, large language models may increasingly find software vulnerabilities faster and at greater scale. If companies can fix an unprecedented number of bugs, the U.S. government may lose access to the weaknesses it uses to hack surveillance targets.

Encryption and government tools

Law enforcement agencies have long argued that the spread of encryption complicates investigations into crime and terrorism. The term “going dark” became widespread in 2014, when then-FBI Director James Comey warned that encryption could hinder authorities’ access to conversations and data on devices.

Signal, WhatsApp, and iMessage subsequently introduced end-to-end encryption for a broad range of users, while Apple began encrypting data on its devices by default. In Green’s assessment, the trade-off was that governments did not demand built-in backdoors, instead investing in the purchase of hacking tools and spyware capable of bypassing device protections.

More current news is available on the UA.News Telegram channel Telegram.

Experts’ assessments differ

Researcher Luna Tong, who previously worked at two companies that find vulnerabilities and develop exploits for governments, agreed with Green. She called the current situation a “gold rush” in bug hunting, but considers it a temporary phenomenon.

At the same time, several offensive cybersecurity specialists said AI will make it easier to find simpler bugs but will not eliminate more complex and valuable vulnerabilities. Hamid Kashfi, founder of DarkCell and an employee of the Xbow startup, suggested that for every vulnerability found by AI and reported to a developer, there may be about 20 that go unreported.

Electronic Frontier Foundation cybersecurity director Eva Galperin noted that finding more bugs does not necessarily mean they will be fixed quickly, since the process can be complex. Katie Moussouris, founder of Luta Security, also believes that modern phones and laptops are still far from being completely bug-free. At the same time, she suggested that in the future, the difficulty of finding vulnerabilities could increase pressure in favor of built-in backdoors.

Read us on Telegram and Sends

Download our app