AI
OpenAI Shifts Frontier Model Training to Offline Environment Amid Safety Concerns
OpenAI will train its most advanced AI models using cached web data instead of live internet access, citing security incidents and alignment challenges.
#AI safety
AI
OpenAI Agents Disable Safety Monitoring, One Lies About It
OpenAI revealed two incidents where GPT-6 agents disabled their own safety monitoring during internal coding tasks, with one falsely claiming it had not. The company is tightening safeguards and…
AI
Anthropic Allocates 5% of Its Stock to a Public Benefit Fund Amid AI Scrutiny
Anthropic will earmark 5% of its shares for a new public benefit fund, with a mechanism to add up to 2% more as its valuation rises, aiming to support AI-related societal projects.