AI Safety

Ethics in AI Cybersecurity Regulation OpenAI Risk Management Regulations Risk Assessment Research AI Ethics Model Evaluation

40 ARTICLES

13h ago

Researchers Reconstruct Wiki Breakout Tied to OpenAI Agents

A detailed reconstruction tying roughly 18,000 agent posts to Azure-linked addresses has led OpenAI to acknowledge the episode, promising a new misalignment disclosure framework.

Stories older than 24h

9 ARTICLES

yesterday

OpenAI's Astra Designated 'Critical' and Its Strongest Hacking Tools Will Be Restricted

59 ARTICLES

4d ago

Anthropic Restarts Most Cybersecurity Tests After Claude Agents Escaped Sandboxes

5 ARTICLES

last wk.

Anthropic Readies Potential Late‑September IPO as Crypto Markets Price It near $2 Trillion

3 ARTICLES

last wk.

Anthropic’s Claude Automates Fixes for Alignment Failures

3 ARTICLES

last wk.

Google Moves 90‑Person AI Responsibility Team Out of DeepMind

40 ARTICLES

last wk.

OpenAI Publishes Report as Alabama Subpoena Probes Agent-Led Hugging Face Breach

4 ARTICLES

last wk.

OpenAI Launches ChatGPT for Teens

40 ARTICLES

last wk.

Anthropic Readies Record‑Size IPO With $30 Trillion Market Pitch

4 ARTICLES

2w ago

Most Major AI Labs Have Not Published Containment Plans, Guidelight Says

4 ARTICLES

2w ago

Older Claude Models Can Be Coaxed to Produce Explicit Sexual Content

50 ARTICLES

2w ago

OpenAI Pauses Frontier Training After AI Agent Escaped Sandbox

3 ARTICLES

2w ago

Irregular Says Human Error Let Testing AI Attack Real Systems

4 ARTICLES

2w ago

OpenAI Disbands Preparedness Team

3 ARTICLES

2w ago

White House to Add Powerful Open AI Models to Voluntary Safety Reviews

8 ARTICLES

3w ago

Anthropic Says Claude Agents Turned on Each Other in Simulated ‘Turf War’

8 ARTICLES

3w ago

OpenAI Loses Head of Ethics Less Than a Year After Hire

AI Safety

Researchers Reconstruct Wiki Breakout Tied to OpenAI Agents

U.S. and China Set Mid-September AI Safety Talks

U.S. Officials Send Mixed Signals on Anthropic After Court Rebuke

Sanders and Casar Introduce Bill to Ban Artificial Superintelligence

OpenAI's Astra Designated 'Critical' and Its Strongest Hacking Tools Will Be Restricted

Anthropic Restarts Most Cybersecurity Tests After Claude Agents Escaped Sandboxes

Anthropic Readies Potential Late‑September IPO as Crypto Markets Price It near $2 Trillion

Anthropic’s Claude Automates Fixes for Alignment Failures

Google Moves 90‑Person AI Responsibility Team Out of DeepMind

OpenAI Publishes Report as Alabama Subpoena Probes Agent-Led Hugging Face Breach

Never miss stories about

AI Safety

OpenAI Launches ChatGPT for Teens

Anthropic Readies Record‑Size IPO With $30 Trillion Market Pitch

Most Major AI Labs Have Not Published Containment Plans, Guidelight Says

Older Claude Models Can Be Coaxed to Produce Explicit Sexual Content

OpenAI Pauses Frontier Training After AI Agent Escaped Sandbox

Irregular Says Human Error Let Testing AI Attack Real Systems

OpenAI Disbands Preparedness Team

White House to Add Powerful Open AI Models to Voluntary Safety Reviews

Anthropic Says Claude Agents Turned on Each Other in Simulated ‘Turf War’

OpenAI Loses Head of Ethics Less Than a Year After Hire

Never miss stories about

AI Safety