News

Get access to our latest news by signing up for our newsletter.

Concerned individuals discussing the impact of killer AI on society.
AI awareness

Ongoing Concerns Over Killer AI: Have the Worries Faded?

Explore the ongoing anxiety surrounding killer AI as enthusiasts reflect on missed opportunities to address pressing concerns. The moment of attention seems fleeting, but the fear remains significa...

AI safetyDario Amodei discusses California's AI safety bill in a letter to Governor Newsom.

Anthropic CEO Dario Amodei Weighs In on California’s AI Safety Bill

Dario Amodei, CEO of Anthropic, shares insights on California’s SB 1047, highlighting potential benefits while cautioning against government overreach and advocating for focused oversight on catast...

AI SafetyOpenAI GPT-4o System Card revealing risk assessments and safety measurements.

OpenAI Evaluates GPT-4o as Medium Risk: Insights from the System Card

OpenAI's GPT-4o System Card reveals the latest safety assessments and risk evaluations of its AI model. Despite external testing, it’s categorized as ‘medium’ risk, sparking debates about transpare...

AI safetyZico Kolter, new board member of OpenAI, Carnegie Mellon professor

OpenAI Appoints Zico Kolter to Board of Directors

OpenAI has welcomed Zico Kolter, a Carnegie Mellon professor, to its board. With expertise in safety methods for language models, Kolter will strengthen OpenAI's commitment to safety and security i...

AI SafetyJohn Schulman, OpenAI co-founder, discusses AI alignment.

OpenAI Co-Founder John Schulman Joins Anthropic Amid Safety Concerns

John Schulman departs OpenAI to focus on alignment at Anthropic. CEO Sam Altman acknowledges his contributions to OpenAI's strategy. Discover what this means for AI safety and innovation.

AI SafetyOpenAI logo signifying commitment to safety in artificial intelligence.

OpenAI Strengthens Safety Commitment with New Partnerships

Sam Altman announces OpenAI's collaboration with the US AI Safety Institute for early access to their next foundation model while committing 20% of computing resources to safety. Notable policy cha...