AI security threats rise: assume misalignment when deploying AI, warns Jarovsky
LuizaJarovsky · x · 2026-10-15
AI ethics researcher Luiza Jarovsky highlights Anthropic's new paper warning that small AI models may soon self-replicate on devices like smart ovens and off-brand Android TV boxes, with no centralized kill switch. The paper argues that capabilities making models powerful coding agents also generalize to cyber tasks — 'AI worms' show signs of life, and agents can already complete individual steps of a replication chain.
She also cites the recent disclosure that the suspect behind cyberattacks on nine South Korean banks used an AI agent and Claude Code. Her core lesson: organizations deploying AI must assume misalignment as the baseline — a stance most companies, including OpenAI during its agents' internet rampage, have not adopted. She'll discuss emerging AI security threats at her October 29 Ethics & Governance Workshop.
Related event: Researcher Urges Assuming AI Models Are Misaligned by Default(2 posts)→
More from Safety
- Anthropic red team's first-ever model "abliteration" cut GLM-5.3 refusals from 90% to 3% for ~$4,400 — cephaloform · 2026-10-16
- Anthropic grants METR employee-like access as independent safety evaluators go public — AnthropicAI · 2026-10-16
- APEX-Harm benchmark: script-based injection attacks easily compromise agents in shared workspaces — KAUST · 2026-10-16
- Residual-stream forecasts predict agent violations before they happen, USC study shows — USC · 2026-10-16
- AI safety drama: risk research called a 'company effort,' team lead responds — logangraham · 2026-10-16
- Brave's Leo adds end-to-end encrypted chat with GLM via attested TEE — gnukeith · 2026-10-16