Leading AI Giants Secretly Collaborate on Self-Regulatory Framework

Key Highlights

• Since July, Anthropic, Google, and OpenAI have conducted regular discussions about establishing a collaborative AI standards organization
• The concept originated from Google DeepMind’s Demis Hassabis, drawing inspiration from Wall Street’s FINRA oversight model
• Sam Altman of OpenAI backs the initiative but emphasizes it must function independently from government involvement
• Opposition has emerged from Meta’s Mark Zuckerberg and several smaller artificial intelligence companies
• An executive order draft from the White House regarding AI regulation was abandoned due to insufficient backing

A coalition of leading artificial intelligence companies is working toward establishing an industry-driven standards organization. According to reporting from The Information, Anthropic, Google, and OpenAI have maintained consistent dialogue since at least the summer months.

ANTHROPIC, OPENAI AND GOOGLE ARE DISCUSSING CREATING AI STANDARDS, WITH THE TALKS HAVING BEGUN BEFORE AMODEI’S ESSAY – THE INFORMATION

— First Squawk (@FirstSquawk) September 13, 2026

The collaborative group convened as recently as the previous week. These discussions continue to progress independent of potential Trump administration participation.

Origins of the Proposal

The initiative was spearheaded by Google DeepMind’s founder, Demis Hassabis. Through an essay published in July, he outlined a vision for a U.S.-centered standards body inspired by FINRA, the regulatory organization that oversees financial markets under SEC supervision.

Hassabis’s framework calls for an entity that would evaluate cutting-edge AI systems prior to public deployment. The proposed structure would operate as a hybrid public-private entity, financed by industry participants but operated by neutral technical specialists.

Anthropic’s CEO, Dario Amodei, reinforced this direction last Saturday through an essay advocating for industry-wide cooperation on evaluation and auditing protocols. These discussions, however, had already commenced well before his public statement.

Leadership Perspectives

During a recent employee town hall, OpenAI’s CEO Sam Altman expressed his endorsement for developing an evaluation and auditing organization for artificial intelligence. He stressed that leading laboratories would need to construct this framework autonomously, without depending on governmental resources.

Altman recognized that the proposal lacks universal industry acceptance. He identified direct cooperation among competing laboratories as the most practical approach available.

Jakub Pachocki, OpenAI’s Chief Scientist, revealed that the organization has been engaging with external entities regarding the establishment of specific standards and promised additional announcements within the next several months.

Industry Opposition

The proposal has encountered resistance. Meta’s CEO, Mark Zuckerberg, voiced apprehensions about a nationwide AI regulatory body during an August telephone conversation with President Trump.

A preliminary White House executive order designed to establish such an organization stalled. The proposal was ultimately discarded after it failed to secure adequate internal administration support before summer concluded.

Speaker of the House Mike Johnson noted the absence of industry-wide agreement on regulatory standards. “There’s no consensus among them,” he stated to CNN.

Cohere’s CEO, Aidan Gomez, voiced disapproval of the proposal on X, cautioning that permitting dominant AI corporations to establish regulations could disadvantage emerging competitors. “Great ideas here from the cartel,” he wrote.

Google, OpenAI, and Anthropic either refused to provide statements or did not reply to CNN’s inquiries

The post Leading AI Giants Secretly Collaborate on Self-Regulatory Framework appeared first on Blockonomi.