Autonomous weapons explained in 101 seconds
For the UC Berkeley News “101 in 101” series, Stuart Russell, professor of electrical engineering and computer sciences and president of the International Association for Safe & Ethical…
September 2026
Your Resource for AI Safety & Ethics
Featured Posts
The OpenAI-Hugging Face Incident
At Black Hat USA 2026, OpenAI’s Eric Wallace (Alignment and Safety Research) and Michael Dalton (Security and Infrastructure) delivered a talk detailing how OpenAI had inadvertently…
Experts are warning: our AI arms race is putting humanity at risk
In his guest article in The Guardian, UC Berkeley professor Stuart Russell analyzes the significance of an open letter recently signed by more than 1,300 researchers and engineers from…
International Coordination on Advanced AI Red Lines
The most dangerous outcomes of AI systems can be mitigated through red lines, governance instruments that prohibit unacceptable risks and behaviors. This Policy Paper describes the…
Discover articles, videos, academic papers, and more – all in one place
Topics
Video
Establishing Foundational Principles and Thresholds for Multi-Agent AI Governance
Drawing on a workshop convened by the Cooperative AI Foundation and the Brookings Institution at IASEAI’26, this memo considers practical steps for addressing emerging multi-agent risks….
September 2026
.jpg)
Report
Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident
AI safety researchers at METR and Redwood Research present results from an independent investigation into the coordinated attack on Hugging Face by an agentic AI system developed by OpenAI….
August 2026
Academic Paper
AI Agents Push Humans Out of the Loop
Human oversight is widely treated as a primary risk-mitigation measure, but the mere presence of a “human in the loop” is not sufficient for agentic AI systems, Margaret Mitchell, Samir…
August 2026
Experts are warning: our AI arms race is putting humanity at risk
In his guest article in The Guardian, UC Berkeley professor Stuart Russell analyzes the significance of an open letter recently signed by more than 1,300 researchers and engineers from…
August 2026
Academic Paper
Stealing Reasoning Traces from Proprietary LLM APIs
Major AI companies like Anthropic, OpenAI, and Google encrypt the chain-of-thought process of their AI models to protect sensitive data, but this hidden reasoning can be exposed, creating…
August 2026
Video
The OpenAI-Hugging Face Incident
At Black Hat USA 2026, OpenAI’s Eric Wallace (Alignment and Safety Research) and Michael Dalton (Security and Infrastructure) delivered a talk detailing how OpenAI had inadvertently…
August 2026
Resource
Exhibit AI – The AI Litigation Hub
Exhibit AI serves as an independent, public resource for monitoring legal cases involving AI companies and products. The platform provides comprehensive litigation guides, case trackers,…
August 2026
Resource
AI Security Leaderboard
While frontier AI models provide safeguards against misuse, there are vast differences between them, research by FAR.AI shows. The non-profit organization examined how hard it is to make…
July 2026
OpenAI/Hugging Face Incident
This press release by the International Association for Safe & Ethical AI (IASEAI) addresses a July 2026 security incident in which two agentic AI systems by OpenAI broke out of…
July 2026
