Google предложила FARO для frontier-ИИ, но защита от обычных рисков под вопросом

Google published the document "A Pragmatic Approach to AI Governance in America" in June 2026, proposing to split AI regulation into two levels: create a new FARO body for the most powerful frontier systems, and leave everyday AI under existing laws.

What FARO Is and Who It Will Cover

Google proposes establishing FARO — the Frontier AI Regulatory Organization — an independent structure to oversee advanced AI systems capable of posing a threat to national security. Under the company's plan, the board will include both independent members and industry representatives. Critics immediately identified the problem: if the industry finances the regulator and occupies seats on its board, the question of genuine independence remains open.

The division into "frontier" and "everyday" AI is the document's key architectural decision. It determines who will fall under new regulation and who will not.

What Remains Outside the New Plan

Everyday AI products — chatbots, automated decision-making systems in employment, finance, and education — Google proposes to regulate through existing legislation: consumer, labor, privacy, and copyright law. In an analysis for Tech Policy Press dated August 3, 2026, researcher Gregory Gondwe argues that this approach underestimates "harm that accumulates gradually — through dependency and repeated decisions."

• Google's document was published in June 2026 under the title "A Pragmatic Approach to AI Governance in America"
• FARO is a new regulator for frontier AI posing a national security threat; the board includes both independent members and industry representatives
• Ordinary AI remains under current legislation without a new regulator
• The principle of regulatory reciprocity may de facto make American standards global

Copyright is a separate pain point. Technically, opt-out tools are available to creators; however, as Gondwe notes, large media companies "have lawyers, engineers, and negotiating capacity" to protect their interests. Independent creators and small publishers lack both the tools to detect violations and the leverage to negotiate.

Who

Pays for Infrastructure and Whose Standards Will Become Global

The large-scale construction of data centers with growing electricity and water consumption raises questions about the distribution of benefits and costs. According to the logic of Google's proposal, infrastructure growth means economic development for regions, but analysts warn: profits concentrate with companies while local communities bear the burden on energy grids and water supply.

The international context complicates the picture. The principle of "regulatory reciprocity" embedded in Google's proposal could automatically extend American standards to the entire world, marginalizing the specific interests of countries in Africa, Latin America, and Asia. Gondwe recommends a number of measures: the majority of seats on regulatory boards should be held by public interest representatives, funding should be independent, audit results should be public, and the movement of officials into jobs at regulated companies should be restricted.

"Large media companies have lawyers, engineers, and negotiating capacity... small creators have no way to either detect a violation or negotiate," writes

Gregory Gondwe, researcher at Tech Policy Press, in an analysis dated August 3, 2026.

What This Means

Google's document sets the framework for the discussion on AI regulation in the US: frontier systems will receive specialized oversight, while a broad class of everyday risks — from algorithmic bias to the copyright of small creators — remains under old legislation. Whether this is sufficient for a rapidly changing market is a question that will define the political debate in 2026–2027.

Frequently Asked Questions

What is FARO?

FARO (Frontier AI Regulatory Organization) is a regulatory body proposed by Google in June 2026 to oversee the most powerful AI systems posing a national security threat. The board will include independent members and industry representatives; critics doubt genuine independence given such a composition and funding source.

How does Google propose to regulate ordinary AI products?

Google proposes to regulate chatbots and automated decision systems through existing legislation — consumer, labor, privacy, and copyright law — without creating new specialized bodies.