{
  "id": 9667656,
  "title": "The companies racing to build frontier AI are now racing to govern it",
  "url": "https://urgent.news/2026/09/25/the-companies-racing-to-build-frontier-ai-are-now-racing-to-govern-it",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-25T00:36:55.000Z",
  "source": {
    "name": "Computerworld",
    "slug": "computerworld",
    "url": "https://www.computerworld.com/article/4226400/the-companies-racing-to-build-frontier-ai-are-now-racing-to-govern-it-2.html"
  },
  "original_language": "en",
  "account": "While AI companies continue to release increasingly capable models, they are also banding together to establish safety standards for frontier AI. The Information reports that Google, OpenAI, and Anthropic are working towards a body called the Standards Authority for Frontier AI (SAFA) that would operate independently of government control. This organization would set guidelines for risk assessment, testing, and pre-release review practices for these powerful AI models, with the goal of officially launching the initiative in early 2027.\n\nNotably, leaders from leading AI companies have called on the United Nations to create safeguards around AI to prevent it from becoming too powerful to control. OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei have urged for such international standards, emphasizing the importance of making AI progress \"safe and beneficial.\" OpenAI's CEO specifically points to the need for advances in alignment research and shared standards, including agreed-upon baselines, common measurements, and incident reporting protocols.\n\nThough enterprises are keeping a close watch on AI safety, their primary concerns lie in how AI will impact their day-to-day operational security. They are worried about potential data exposure, platform vulnerabilities, hallucinations affecting corporate workflows, and AI compromising compliance efforts. Enterprises want to know the return on investment (ROI), additional risks, new capabilities, risk-reward balance, and whether the technology vendor is responsive to their needs.\n\nIndependent technology analyst Carmi Levy suggests that enterprises should proactively minimize their AI-centric risk profiles while maximizing their ability to derive value from AI. This can be achieved by enforcing standards for vendors, requiring safety and security datasheets for new models, extending existing change management processes to AI models, and establishing internal AI proof of concept or governance committees. Additionally, enterprises should consider data privacy concerns, IP leaks, and mandate AI literacy training to ensure users understand how to spot and verify AI-generated output.",
  "summary": "Even as they continue to release ever more capable competing models in a regular cadence, the top AI companies are joining forces to set AI safety standards. According to The Information , Google, OpenAI, and Anthropic are reportedly working together to create a body tentatively called the Standards Authority for Frontier AI (SAFA). It would operate independently of government control, and set…",
  "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."
}