Anthropic Report: AI Agents Automated a Full Cloud Breach, 1.8M Apps Scanned
Comfortable_Gene5180 · reddit · 2026-09-21
Anthropic's latest threat intelligence report describes a case where criminals used AI agents to automate much of a cyberattack — from finding exposed credentials to breaching systems and exfiltrating data — an approach the company calls "vibe hacking."
Key details:
- Attackers scanned 1.8 million Android apps for exposed credentials and keys
- AI automated reconnaissance, scripting, and data discovery
- Software suppliers were targeted to reach their customers' data
- Stolen AI/API keys were used to run further activity on victims' own compute
- In one case, over 2,100 login tokens across 40+ company accounts were collected in roughly 34 hours
The takeaway: security controls designed around human-speed attackers may not hold up when parts of an attack can be automated.
More from Safety
- Senate Bill Would Force Big Platforms to Accept AI Agents, Raising Auth Privacy Questions — Prof_CarlaReyes · 2026-11-12
- United Airlines Bans AI Agents From Booking Flights Over Data and Legal Risks — rohanpaul_ai · 2026-11-12
- Stanford HAI blueprint: 56 US agencies report 3,611 AI use cases, evaluation lags — StanfordHAI · 2026-11-12
- Focused task framing boosts agent boundary testing by ~60 points; VM escape still fails — perplexity_ai · 2026-11-12
- Perplexity red-teams its SPACE VM sandbox: CTF prompts mislead agent security testing — perplexity_ai · 2026-11-12
- Narayanan: blanket bans on teen AI companions are bad policy; incentivize pro-social design instead — random_walker · 2026-11-12