What's Happening?
Aidan Gomez, co-founder and CEO of Cohere, has publicly criticized a proposed AI standards body, labeling it a 'cartel.' In a September 13 essay, Gomez accused 'market-dominant AI companies from Silicon Valley,' specifically naming Anthropic CEO Dario
Amodei and implicitly including OpenAI and Google, of seeking an 'antitrust waiver' to legitimize their coordinated efforts. He argues that a safety regime designed by a few leading labs would entrench their current advantages and force other AI developers to conform to their established standards. Gomez highlighted historical parallels, citing the SEC's designation of three firms as Nationally Recognized Statistical Rating Organizations in 1975, which he claims led to the rating of subprime mortgage securities as triple-A, and the 1985 Motor Vehicle Block Exemption in Europe, which protected incumbents. He contends that such initiatives, while ostensibly for safety, ultimately serve to limit competition and protect the market position of the dominant players.
Why It's Important?
This development is significant for the U.S. technology industry and regulatory landscape, particularly concerning artificial intelligence. Gomez's strong critique brings to the forefront the debate over who should govern AI development and the potential for anti-competitive practices within a rapidly evolving sector. If dominant AI companies are allowed to set industry standards without broad input, it could stifle innovation from smaller firms and startups, limiting the diversity of AI applications and potentially concentrating power in the hands of a few. This could have profound implications for national competitiveness in AI, as well as for consumer choice and data privacy. The call for an antitrust waiver also raises concerns about regulatory oversight and the potential for powerful corporations to influence policy in their favor, impacting the future direction of AI governance and market structure.
What's Next?
Gomez proposes an alternative framework for AI governance, advocating for four pillars: an evidence-based, internationally published risk framework funded by public research bodies; mandatory developer transparency through model cards and incident reporting; independent testing focused on 'genuinely dangerous capabilities' like cyberattacks and bioweapons; and independent assurance mechanisms modeled after financial, aviation, and nuclear regulations. This counter-proposal suggests a push for a more decentralized and transparent approach to AI standards. The ongoing discussions among Anthropic, OpenAI, and Google, as reported by AI Weekly, indicate that the formation of an AI standards body is actively being pursued. The coming months will likely see continued debate among industry leaders, policymakers, and civil society groups regarding the most effective and equitable way to regulate AI, balancing innovation with safety and competition. The outcome could shape the future regulatory environment for AI in the U.S. and globally.
Beyond the Headlines
The controversy surrounding AI standards touches upon fundamental questions of power, ethics, and the future of technological development. Gomez's use of historical analogies, particularly the subprime mortgage crisis, underscores the potential for self-regulation by powerful entities to lead to systemic risks and market failures. The debate highlights the tension between rapid technological advancement and the need for robust governance structures that protect public interest and foster fair competition. Beyond economic implications, the control over AI standards could also influence the ethical development of AI, including issues of bias, accountability, and the societal impact of autonomous systems. The outcome of this debate will not only determine the commercial landscape of AI but also its broader role in shaping society, potentially setting precedents for how emerging technologies are regulated in an increasingly complex world.
