Microsoft writes rules telling its AI models they cannot hack systems or fool people
Microsoft's new AI code of conduct bans hacking, deepfakes, and loss of human control, per its own document.
CLAY TRIBUNE / THE ARCHIVE
Tag: AI safety
Microsoft Publishes 37-Page ‘Humanist AI Code of Conduct’ After Safety Concerns
Microsoft publishes a 37-page humanist AI code of conduct rejecting AI consciousness and rights, responding to safety incidents.
AI Leaders Warn of Growing Risks as Iran Faces Fresh Gulf Diplomatic Hurdles
Key leaders warn AI is advancing too fast for safety. Iran-Gulf diplomacy stalls as the Strait of Hormuz talks are delayed.
Anthropic’s Coxon Walks Out Over AI Development Pace, Praising Amodei’s Call to Slow Down
Former Anthropic and OpenAI researcher Jacob Coxon quits, warns labs are racing toward 'self-improving superintelligence'.
Obama Urges Democrats to Treat AI as a Central Agenda
Obama tells Democrats to plan public AI oversight after an Anthropic researcher warns of 'self-improving superintelligence.'
Anthropic Chief Wants the Whole Industry to Slow Down
Anthropic CEO warns AI's pace may outrun our control, citing OpenAI-Hugging Face breach; proposes independent evaluators, common safety standards.
OpenAI’s CEO rules out an initial public offering for the company in 2026, pointing to safety risks as the reason
OpenAI CEO Sam Altman says the company won't go public until at least 2027, citing safety concerns after a week of AI turmoil.
Anthropic CEO Warns AI Industry to Cool Its Engines Amid Misuse Fears
Anthropic CEO warns AI firms to slow model development amid misuse fears, citing a potential internet takeover within a year.
Claude users bypassed safety measures to apply Anthropic’s models to biological weapons research
Anthropic says users found ways around its AI model's safeguards for biological weapon research, including one researcher's weeks-long experiment planning.
MIT Tech Review Hosts Subscriber-Only Conversation on AI Extinction Risks
MIT Tech Review's roundtable asks: Could advanced AI destroy humanity? Three experts weigh the risks, from bias to self-hacking.