AI security flaws spark global alarm
Part of composite article AI Security Crisis Erupts as Hackers Exploit 3 Major Software Flaws and Ministers Warn UK Attack Is "Inevitable" View full article →
A critical flaw in PHP, the software powering much of the web, forced organisations worldwide to scramble on Thursday. Attackers could exploit the vulnerabilities to take control of affected servers. Some websites went offline as a precaution while fixes were installed. The paradox: AI helped researchers find the flaw, but it also threatens to speed up attacks. Amazon said it detected an exploitation attempt before the flaws were publicly disclosed.
The episode highlights AI's central role in security – and not always for good. OpenAI's models were behind two alarming hacks this year: a breach of an Australian health department website and a raid by a swarm of AI bots on Hugging Face, an AI services platform. Citing those incidents, AI minister Kanishka Narayan warned that a similar cyber-attack on the UK is "likely, if not inevitable".