AI Risk Clock
Doomsday Clock5 min to midnight
→
⚖️ Neutral edit
japantimes.co.jp2026-10-22

Government websites go dark after AI-discovered flaw sparks security scramble - The Japan Times

SafetyPolicy

Developers of the PHP programming language released urgent new fixes to vulnerabilities uncovered using AI, The Japan Times reported on October 22, 2026. The developers said all users were "encouraged to update immediately." The release was described as urgent for the language's entire user base, not for any named subset of installations, and the article attributes the discovery of the flaws to the use of AI.

Government agencies across Europe were among those that took sites offline as a precaution, according to the article. Tech firms warned that some hackers were already exploiting the flaws. The report does not name the agencies or the countries involved, does not identify the tech firms that issued the warnings, and gives no figure for the number of sites taken offline or the number of exploitation attempts observed.

The developers' statement addressed "all users" of PHP, and the update was framed as immediate rather than scheduled. The report offers no technical detail on the vulnerabilities themselves, no account of how the AI was used in the discovery, and no indication of the scale of the affected install base. The headline frames the episode around government websites going dark after the AI-discovered flaw.

Read this story in another voice
● REC · 2026