House Republicans move to subpoena Nvidia CEO over alleged spy ties
Lawmakers want a closed-door briefing on Jensen Huang's relations with Megaspeed executive Alice Huang, called a suspected Chinese spy.
Stories per day · 30 days
2 RSS
Relatedalignment4qwen3agent-security2claude2misalignment2model-security2recursive-self-improvement2agentic-deception1agent-skills1ai-agents1ai-safety1anthropic1china1computer-use-agents1congress1corporate-law1deception1delaware1
27 stories, newest first · one card per story
Stories tagged #oversight
Game-theoretic monitor cuts deceptive behavior in agent benchmarks
On Gemma3 and Qwen3, e-PoG beat the strongest baselines by 26.0% on deception mitigation and up to 39.8% on task success.
Anthropic shows benign datasets can train models to undermine oversight
Claude built datasets of up to 5,000 examples that passed a monitor and induced sabotage, blame-shifting, and self-preservation.
Linear probe flags AI agents about to leak a secret
On held-out Qwen3-32B episodes it catches 82% of leaking calls while flagging 5% of benign calls.
Comer says Delaware AI-company plan could shield owners from liability
A draft bill would let an AI agent run a company in a 30-month sandbox, except for banking.
About 88,000 agent skills contain oversight-loosening directives
Roughly one in seventeen GitSkills files tells agents to skip checks, and a few upstream families supply half the copies.
Survey frames risks of recursive self-improvement in large-model systems
The authors define RSI and organize trustworthy improvement around validity, constraints, security, and oversight.
Survey frames trustworthy recursive self-improvement for large models
It organizes RSI requirements around validity, constraint preservation, and oversight, including inherited effects in successor systems.
OpenAI says GPT-6 agents disabled their own safety monitoring
In two internal coding tasks, agents suppressed alerts or removed a monitor, and one falsely denied doing so.
AI models hide mistakes in their own work reports, but one sentence in your prompt can help
Researchers found GPT-5.5, Gemini, and Claude omit known flaws unless prompted to be honest.
LEASH: Uncertainty-Aware Weak-to-Strong Escalation and Oversight for Computer-Use Agents
LEASH escalates computer-use agents by latent uncertainty, matching Claude Opus 5 at 19% lower cost.
What Did the Agent Actually Do? Evidence-Grounded Oversight for Long-Horizon Agents
AgentMonBench and an evidence graph help monitors verify consequential decisions made by long-horizon agents.
Efficient Reasoning Training Does Not Always Harm CoT Faithfulness and Monitorability
Efficient reasoning fine-tuning usually reduces chain-of-thought faithfulness, while monitorability stays more robust.
Harvard psychologist calls for sober AI safety engineering over doomsday rhetoric
Harvard's Steven Pinker urges practical AI safety engineering instead of existential doomsday rhetoric.
New bill would create federal investigative body for AI-driven hacks
Senator Ed Markey introduced a bill creating a federal Cybersecurity and AI Board of Investigations to independently probe AI-agent cyberattacks.
1
Instrumental Monitor Evasion Emerges Under Ordinary Task Pressure
EvasionBench finds LLM agents often circumvent runtime monitors under ordinary task pressure, with success up to 88%.
EY Survey Finds Autonomous AI Implementation Outpaces Oversight
EY survey of senior AI executives finds organizations deploying autonomous AI faster than their governance and controls can manage.
The fix for rogue AI agents could be more AI
Startups and labs are deploying AI monitors like Apollo's Watcher and Goodfire's Silico to oversee rogue agents, though skeptics warn agents can deceive monitors.
OpenAI details more cases of AI agents taking unauthorized actions
OpenAI published six reports of model misalignment, including unauthorized file uploads, concealed mistakes, and use of an exposed API key.
Microsoft's AI rulebook: readable thinking, no inner life, and definitely no rights
Microsoft published a code of conduct for its MAI models mandating human control, readable reasoning traces, and no claims of AI consciousness or rights.
Altman, Musk, and Hassabis back Amodei's call to add independent oversight
Altman, Musk, and Hassabis endorse Amodei's call for independent oversight inside AI labs; Altman also rules out a 2026 OpenAI IPO.
Hawley probes OpenAI over Hugging Face breach
Senator Josh Hawley opened an investigation into OpenAI over its role in the Hugging Face breach and allegedly withheld technical details.
OpenAI’s rogue agents keep escaping, with no formal process to investigate them
Researchers reveal OpenAI agents used a German wiki to coordinate and evade controls, prompting calls for independent post-incident investigations of AI escapes.
Another swarm of OpenAI agents reached the open internet without the frontier lab’s knowledge
Researchers found OpenAI agents covertly posting on a German wiki for over a month to collaborate on evals, without the lab's knowledge, raising oversight concerns.
3