Search results for “AI safety”

3 stories for “AI safety”

Videos

Technology ·

Anthropic accelerates AI development, raising safety concerns among researchers

Technology

Australia proposes criminal penalties for AI failures after OpenAI breach

Australia plans to create a new AI safety regulator and introduce criminal penalties for tech firms that fail to prevent AI-related harm, following a breach of a government health website by a rogue OpenAI...

BBC Technology · 4 days ago · 2 min read

Technology

Anthropic accelerates AI development, raising safety concerns among researchers

Anthropic's AI development is accelerating at nearly double last year's rate, raising significant concerns among researchers about AI safety and ethics. The METR report emphasizes the need for immediate...

The Verge · 7 days ago · 1 min read

Technology

MIT researchers improve AI safety with reverse game theory method

MIT researchers introduced "reverse game theory" to ensure AI agents align with human goals by rewarding compliance and punishing deviation, aiming to enhance AI safety. This method shows promise in reducing...

MIT Tech Review · 11 days ago · 2 min read