AI-powered cyberattacks surge as developers expose web vulnerability
Web developers warn that AI-driven attacks are rapidly targeting sites after a bug was discovered, with Cloudflare seeing spikes, Microsoft citing a China-linked group, and chatbots used against banks.
After developers used AI to locate a vulnerability affecting most websites, Cloudflare reported a surge in exploitation attempts. Microsoft said a China-linked hacking group leveraged AI to breach an IT services firm, and commercially available chatbots have been employed in bank attacks. Experts note that OpenAI's "defender's window" is shrinking as AI tools become more accessible to attackers.
Why it matters
AI tools are enabling faster, broader cyberattacks, threatening the security of corporate and government systems worldwide.
How this story developed
- Oct 12 Cyberattack disrupts refrigeration systems at Intermarché and Coopérative U stores
- Oct 12 The attack specifically exploits outdated Danfoss software versions older than version 4.
- Oct 14 Additional reports note that stores of Monoprix, Leclerc and Carrefour also experienced temperature overrides.
- Oct 21 PHP maintainers announced they will release emergency updates for two high-severity flaws, each rated 9.2/10, in just over a day.
- Oct 21 EU’s CERT‑EU recommended that organisations unable to patch consider taking vulnerable sites offline.
- Oct 21 AI tools identified two severe remote-code execution bugs in PHP, forcing governments and companies to pull or patch sites as a fix was rushed out.
- Oct 22 German state portals, including Berlin and Hessen, were restored after applying the emergency PHP patches.
- Oct 22 Emergency patches for PHP 8.x have been released as the vulnerability spreads.
- Oct 22 Additional governments—including Denmark, the United Kingdom, Hungary, France, the Netherlands and Finland—temporarily disabled public websites while applying the PHP patches.
In this story
Related stories
16 in this thread