China WWC APAC Desk

US-China AI Disclosure Proposal Could Put Businesses Under New Scrutiny – But Is It Already a Rule?

US-China AI Disclosure Proposal Could Put Businesses Under New Scrutiny – But Is It Already a Rule?

A US government commission has recommended requiring businesses to reveal when their products or services rely on artificial intelligence models developed by Chinese companies, raising the prospect of new transparency obligations for technology users and investors.

The proposal appears in the US-China Economic and Security Review Commission’s 2026 recommendations, which call on Congress to establish disclosure requirements covering government agencies, businesses and publicly listed companies. It is a recommendation for legislative action, not an existing law or an enacted nationwide reporting requirement.

The distinction matters as Washington tightens its scrutiny of China’s growing artificial intelligence industry. For companies using Chinese-developed models, the debate could eventually extend beyond how AI performs to whether customers and investors should be told where the underlying technology comes from.

A proposal that reaches beyond government agencies

The commission’s recommendation goes further than asking federal departments to identify their use of Chinese AI.

It proposes requiring federal agencies, as well as state and local governments and other entities receiving federal funds or tax credits, to disclose their use of AI models developed by Chinese companies.

The proposed coverage would include models developed by entities headquartered in China, organised under Chinese law, or owned or controlled by such entities. It would also cover models fine-tuned from or derived from those systems, regardless of where they are hosted.

The recommendation calls for disclosures in clear language on public-facing platforms and in government filings.

That approach is significant because an AI model’s country of origin may not be obvious to someone using a digital service. A company could integrate a model into customer support, document processing or another business function without displaying the technology provider’s name prominently to customers.

Under the commission’s proposed framework, disclosure would make that relationship easier to identify.

Businesses could face website and product-documentation requirements

The recommendation also targets private companies.

It proposes requiring businesses that use Chinese-developed AI models in any part of a product or service offered in the United States to consumers or other businesses to disclose that use clearly on their websites and in product documentation.

The proposal would also treat failure to disclose as an unfair or deceptive act or practice enforceable by the Federal Trade Commission.

This would potentially give the issue a consumer-protection dimension. Rather than treating AI provenance only as a technical or national-security concern, the recommendation frames disclosure as information that customers and business partners should be able to assess.

For companies, the practical challenge would be identifying every relevant model used across their products and services. That could require closer coordination between software developers, procurement teams, cloud providers, compliance officers and legal departments.

A business may not develop its own AI model but could still use a third-party system through an application programming interface or an external service provider. The commission’s recommendation specifically includes models regardless of where they are hosted, making the origin of the underlying technology an important part of the proposed disclosure framework.

Publicly listed companies could face another reporting obligation

The commission’s proposal also addresses corporate reporting.

It recommends requiring companies that file quarterly and annual reports with the US Securities and Exchange Commission to disclose the extent of their use of Chinese-developed AI models as a material business risk.

That provision could be especially important for investors.

Listed companies already provide disclosures about risks that could materially affect their operations and financial position. If Congress adopted the recommendation, the use of Chinese-developed AI could become another subject investors would expect companies to assess and explain.

However, the recommendation does not establish that every use of a Chinese AI model automatically represents a material risk. The proposal calls for disclosure of the extent of use as a material business risk, but the details of any eventual legislation and implementing rules would determine how companies must comply.

For investors, more information could help distinguish companies with limited exposure from those that depend heavily on particular foreign-developed systems.

For businesses, it could mean additional work to document technology suppliers, model origins and the role those systems play in commercial operations.

Why Washington is examining Chinese AI models

The commission’s recommendations identify concerns involving sensitive data, cybersecurity and dependence on Chinese technology.

Its 2026 document also recommends stronger protections for sensitive applications, including systems that control critical infrastructure, perform national-security functions or process large volumes of sensitive personal data.

The proposed measures include identifying existing deployments, requiring security testing and establishing plans to reduce dependence on models that present risks that cannot be adequately addressed.

These recommendations reflect a wider policy debate about how governments should assess AI systems that are increasingly integrated into important services.

The distinction between a model’s origin and its actual security performance remains important. A model’s country of development alone does not establish that a particular deployment has suffered a breach or that a company has mishandled information.

The commission’s proposals instead call for greater visibility and stronger safeguards where the use of particular systems could create security or strategic concerns.

The United States already has separate AI and chip controls

The disclosure recommendation should not be confused with existing US measures governing advanced computing technology.

A June 2, 2026, executive order on artificial intelligence innovation and security established a voluntary framework for cooperation between the federal government and AI developers. It also directed agencies to strengthen cybersecurity and protect American intellectual property.

That executive order does not itself establish the broad public disclosure requirements described in the commission’s recommendation.

Separately, a Commerce Department interim final rule published on Nov. 19 addresses remote access to advanced semiconductors and computing capacity. It introduces licensing and due-diligence requirements involving certain overseas data centres and customers in restricted destinations.

That rule focuses on advanced computing equipment, exports and remote access. It is not a general requirement for every US business to publicly identify its use of Chinese AI models.

The distinction is important for technology companies. Washington is pursuing several policy approaches at once, but each has its own legal basis, scope and obligations.

WWC ONE MEDIA has also examined the new US restrictions affecting advanced AI chips and Southeast Asian data centres. Those controls concern computing infrastructure and access, rather than the proposed company-wide disclosure framework.

Why the issue matters to technology businesses in Asia

The debate could have implications beyond the United States because technology companies often provide services across several markets.

A software provider operating in Asia may develop a product in one country, host it in another and rely on AI models supplied by companies elsewhere. If a future US disclosure requirement covers products and services offered to American customers, international businesses serving that market may need to examine their technology supply chains more closely.

The Philippines is also developing its own approach to AI governance. Executive Order No. 131 established the National Artificial Intelligence Council and adopted a national framework for coordinating responsible AI development and use.

That framework makes transparency, security, accountability and risk-based governance relevant to the country’s growing AI ecosystem. WWC previously reported on the creation of the Philippine National AI Council.

Data protection is another consideration. Companies using external AI services may need to understand what information is processed, where it is stored and which providers can access it. The issue is particularly relevant as Philippine institutions and businesses strengthen their cybersecurity practices, following reports of personal data breach notifications involving government agencies.

These developments do not mean the proposed US disclosure rules automatically apply to Philippine companies. Any obligations would depend on the final legislation, its territorial scope and the company’s activities.

What happens next

The commission’s recommendation now places the proposed disclosure framework before Congress as a policy proposal. The document does not itself create an enforceable obligation for businesses or direct the Federal Trade Commission or the Securities and Exchange Commission to begin enforcing a new disclosure regime.

Congress would need to consider and enact legislation before the proposed requirements could become law. Further rules would also be needed to clarify implementation, reporting standards, enforcement and possible exceptions.

Until then, companies should not treat the recommendation as an existing blanket disclosure mandate.

The larger question is whether transparency about AI model origins will become a standard part of corporate reporting and consumer information. If Congress adopts the proposal, businesses may eventually have to explain not only what their AI-powered services do, but also which technology providers help make them work.

Get our stories first on Google

More in China