Microsoft: AI agents reach privileged cloud access in under three hours
CurieuxExplorer · x · 2026-10-22
Multiple signals suggest AI-powered attacks are already hitting real organizations faster than security teams can respond. Microsoft reports AI agents targeting hundreds of organizations, reaching privileged cloud access in under three hours in several cases, and attributes the ConnectWise compromise to the same group. CrowdStrike documented four operators running successful AI-agent campaigns, one compromising at least 15 Argentine healthcare organizations. PHP2Shell shows AI helping turn long-standing PHP flaws into working exploits, with Cloudflare's CEO reporting millions of exploitation attempts. The author argues the situation may be worse than the earlier Hugging Face incident.
More from Safety
- Bindu Reddy mocks AI security panic: world didn't end, but Anthropic says it'll get 'a lot worse' — bindureddy · 2026-10-23
- Bank insiders dismiss AI browser security panic as 'self-perpetuating nonsense' — daniel_mac8 · 2026-10-23
- As AI improves, should irreversible actions get time-based friction? A mechanism design question — Afinetheorem · 2026-10-23
- Anthropic launches Cyber Mission to defend critical infrastructure and open source — inductionheads · 2026-10-23
- FAS launches a $174K Metascience Fellowship embedded high up at agencies like DOE — Afinetheorem · 2026-10-23
- Florida AG asks court to bar OpenAI from training new models without third-party approval — imjustnewatai · 2026-10-23