Google, OpenAI, Anthropic launch self-regulatory AI safety body SAFA
Google, OpenAI, and Anthropic are advancing plans to establish the Standards Authority for Frontier AI (SAFA), a self-regulatory body to set safety and testing standards for advanced AI models. The initiative, targeting launch by end of 2025 or early 2027, follows stalled U.S. federal AI regulation. SAFA would support third-party pre-deployment testing and incident reporting. Leadership candidates include Sriram Krishnan, Arati Prabhakar, Condoleezza Rice, and David Friedberg.
IllustrationEditorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page reads the event directly, while its address stays stable when the title changes.
- Summary covers the current reports
Cross-source coverage
Reporting timeline
Google, OpenAI, Anthropic Reportedly Plan to Create AI Safety Standards Body
According to a report by Caixin citing sources, Google, OpenAI, and Anthropic are advancing plans to establish an AI safety standards organization tentatively named the 'Standards Authority for Frontier AI' (SAFA). The initiative aims to operate independently without government oversight, focusing on concrete testing and audit systems rather than broad ethical guidelines. Specific measures include supporting third-party safety testing before model deployment and establishing incident reporting rules. The companies have reportedly approached Sriram Krishnan, a former White House AI policy advisor, for the CEO role. The effort follows stalled federal AI regulation attempts in the U.S., including a shelved White House executive order. The idea originated from Google DeepMind CEO Demis Hassabis in July, who proposed a model similar to the U.S. Financial Industry Regulatory Authority (FINRA). However, analysts caution that the body's success depends on gaining broad industry support and clarifying its relationship with existing regulators. Critics also worry that self-regulation by leading labs could entrench the dominance of major companies and raise compliance barriers for smaller AI developers.
Read sourceGoogle, OpenAI, Anthropic Plan AI Safety Standards Body Without Government Oversight
According to a report by The Information citing sources, Google, OpenAI, and Anthropic are advancing plans to establish an artificial intelligence safety standards organization that would operate without government oversight. The self-regulatory body, tentatively named the 'Standards Authority for Frontier AI,' aims to launch by the end of 2025 or early 2027. The working group has approached several prominent individuals to serve as CEO, including former venture capitalist and Trump administration AI policy advisor Sriram Krishnan, as well as former Biden administration official Arati Prabhakar. Potential chair candidates include former U.S. Secretary of State Condoleezza Rice and entrepreneur David Friedberg, though some candidates said they have not yet been contacted. The organization would support third-party testing of AI models before deployment and establish norms for how developers report safety and security incidents. The group is still discussing whether to conduct its own AI model safety and capability tests. The report notes that the White House has told the companies they must first reach industry consensus on standards before the government will get involved.
Read sourceGoogle, OpenAI, Anthropic launch SAFA self-regulatory body; Condi Rice, Friedberg, Krishnan considered for leadership
In an exclusive report, the Standards Authority for Frontier AI (SAFA) is being launched as a self-regulatory body by Google, OpenAI, and Anthropic. The report indicates that the companies have considered former U.S. National Security Advisor Condoleezza Rice and entrepreneur David Friedberg for the role of chair, and venture capitalist Sriram Krishnan as CEO. The initiative represents a major industry effort to establish voluntary standards and governance for frontier AI development, though specific details on the body's mandate, funding, and operational scope have not been disclosed in this initial announcement.
Read sourceShow 3 older updatesHide older updates
Google, OpenAI, and Anthropic Push to Form AI Safety Standards Body by Year-End
According to a report by The Information on September 24, Google, OpenAI, and Anthropic are advancing plans to establish a self-regulatory organization called the 'Standards Authority for Frontier AI' (SAFA), with a target launch by the end of 2024 or early 2027. The initiative aims to create unified safety and security standards for companies developing the most advanced AI models. The push comes after efforts to create a joint government-industry AI regulatory framework stalled, partly due to opposition from Meta CEO Mark Zuckerberg and Nvidia CEO Jensen Huang. The White House reportedly told the companies to reach internal consensus before further government action. SAFA would focus on setting standards for pre-deployment third-party testing, incident reporting, and auditor certification. The group is discussing leadership candidates including former Trump AI advisor Sriram Krishnan and former White House science chief Arati Prabhakar. Critics worry SAFA could create barriers for open-source developers and may not effectively address AI governance challenges.
Read sourceGoogle, OpenAI, and Anthropic Form New AI Safety Standards Organization to Fill Regulatory Gaps
According to a report from The Information, Google, OpenAI, and Anthropic are jointly establishing a new artificial intelligence safety standards organization, tentatively named the 'Frontier AI Standards Organization.' The three companies plan for the organization to operate independently, aiming to fill gaps in government regulation of AI. The initiative is still taking shape, and further details about its structure and specific standards are expected to emerge as it develops.
Read sourceGoogle, OpenAI, and Anthropic Form New AI Safety Standards Organization
According to a report from The Information, Google, OpenAI, and Anthropic are forming a new artificial intelligence safety standards organization tentatively named the 'Frontier AI Standards Organization.' The three companies plan for the body to operate independently, aiming to fill gaps in government regulation of AI. The initiative comes as governments worldwide are still developing frameworks for overseeing advanced AI systems, and the new organization is expected to establish voluntary safety standards for frontier AI models. The report indicates that the organization is still in its formative stages, with details on its structure and membership yet to be finalized.