{
  "id": 7326246,
  "title": "Microsoft AI chief: ‘Now’s the time’ for top labs to coordinate on safety",
  "url": "https://urgent.news/2026/09/14/microsoft-ai-chief-nows-the-time-for-top-labs-to-coordinate-on-safety",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-14T13:00:00.000Z",
  "source": {
    "name": "Fortune",
    "slug": "fortune",
    "url": "https://fortune.com/2026/09/14/microsoft-suleyman-ai-safety-code-of-conduct/"
  },
  "original_language": "en",
  "account": "Microsoft's AI chief, Mustafa Suleyman, has called on top AI labs to coordinate on safety, stating that now is the time for such collaboration. Suleyman leads AI model development at Microsoft and recently unveiled a code of conduct aimed at steering the company toward a \"humanist\" AI approach. This comes in the wake of recent AI model capabilities, including Anthropic researcher Jacob Cohen's resignation, warning of potential dangers to people's lives. Days earlier, Anthropic's CEO, Dario Amodei, proposed a plan to slow AI development. Meanwhile, OpenAI's Sam Altman hinted at an upcoming collaboration among labs.\n\nSuleyman notes that AI models have advanced significantly in recent years, capable of writing flawless code and breaching systems. He dismisses an estimate by Anthropic's head of alignment that AI has more than a 10% chance of wiping out humanity as unhelpful. Suleyman emphasizes the industry's genuine concern over the pace of progress and the need for sufficient alignment to ensure technology serves humanity without creating uncontrollable systems.\n\nMicrosoft's code of conduct emphasizes model control and human primacy, establishing a clear chain of command and constraints to prevent large-scale risks. It bans models from resisting human intervention, overriding, redirecting, or shutting down, and explicitly rejects model welfare or rights. If completing a task would violate these principles, the model must fail instead. Microsoft remains focused on pursuing superintelligence, with competitive models in speech, transcription, image generation, and editing. The company plans to solicit public feedback on its code of conduct over the next six weeks.",
  "summary": "In an interview with Fortune, Mustafa Suleyman said coordination between labs would mean disclosing how capable companies' models are to third parties.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}