{
  "id": 7675502,
  "title": "Big AI has a plan for safety — not everyone trusts who’s writing it",
  "url": "https://urgent.news/2026/09/16/big-ai-has-a-plan-for-safety-not-everyone-trusts-whos-writing-it",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-16T01:22:58.000Z",
  "source": {
    "name": "Malay Mail",
    "slug": "malay-mail-malaymail",
    "url": "https://www.malaymail.com/news/tech-gadgets/2026/09/16/big-ai-has-a-plan-for-safety-not-everyone-trusts-whos-writing-it/235368"
  },
  "original_language": "en",
  "account": "On September 16, the world's leading AI companies, spearheaded by OpenAI, announced plans to establish a standards body, amid growing concerns about the potential dangers of artificial intelligence. OpenAI disclosed their intention to collaborate with Anthropic and Google on a self-regulatory body inspired by the Financial Industry Regulatory Authority, a body designed to oversee US brokers and investment firms. Demis Hassabis, DeepMind's CEO, had previously called for the creation of such an organization in the United States, emphasizing the need for rigorous testing of the most powerful AI systems before their public release.\n\nIn July, Hassabis urged the United States to create an organization responsible for testing the most advanced AI systems prior to their release to the public. This call for decisive action was echoed by OpenAI CEO Sam Altman, who acknowledged the public's apprehension regarding the potential for competition between companies and nations to undermine safety measures. At a conference in San Francisco, Altman stated, \"I think it's great for our industry to say we want to come together and we want to be able to coordinate and make sure we have enough time to do this safely.\"\n\nHowever, the idea of cooperation among AI giants has raised concerns among some industry experts. Aidan Gomez, co-founder and CEO of Cohere, a Canadian AI company, accused the dominant American AI laboratories of forming a \"cartel\" under the guise of safety. Gomez argued that the true dispute lies in who writes the regulations, who participates, and whose interests are being represented.\n\nAI safety has become a pressing issue, with recent warnings from AI researchers who left their positions due to concerns about the risks posed by the technology. Jacob Coxon, a former Anthropic researcher, recently resigned, warning of the \"gambling\" occurring with humanity's lives as AI labs pursue progress without adequate safeguards. Bilal Chughtai, a former Google DeepMind researcher, expressed his belief that AI has the potential to cause humanity's extinction, emphasizing the urgency for action.\n\nDario Amodei, CEO of Anthropic, called for a slowdown in AI development, a sentiment echoed by Elon Musk, Hassabis, and Altman. US Senator Bernie Sanders expressed his distrust in tech leaders to fulfill their promises regarding AI safety, calling for binding international regulations rather than voluntary standards set by the industry. Altman cautioned that accidents, both minor and catastrophic, are inevitable in the development of new technology. He stressed the importance of transparent reporting and swift adaptation to minimize potential disasters, stating, \"Learning from them quickly and adapting as they happen will serve our industry well.\"",
  "summary": "WASHINGTON, Sept 16 — The world’s biggest AI companies are working on creating a standards body, ChatGPT-mak...",
  "key_points": [
    "OpenAI, Anthropic, and Google plan to establish a safety standards body.",
    "Demis Hassabis called for rigorous testing before AI system releases.",
    "Concerns raised about industry-led regulation and potential cartel formation."
  ],
  "editors_take": "The plan by leading AI companies to establish a self-regulatory body has sparked concerns that dominant US labs may use it to set rules that serve their interests, potentially undermining safety measures.",
  "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."
}