Blab AI
Senate Draft Bill Aims to Vet AI Agent Providers for Consumer Safety
On June 29 2026, U.S. Senator Mark Warner (D‑Va.) released a discussion draft of the Artificial Intelligence Access, Gatekeeper Exchange, and Nondiscriminatory Transfer (AI AGENT) Act, a proposal that would create a federally vetted list of AI agent software providers for use on large online platforms.
Under the bill, the Federal Trade Commission (FTC) would authorize independent bodies to evaluate AI agent vendors. These bodies would assess whether a product satisfies baseline protections for privacy, data security, and the user’s interests. Each AI agent would also need to link to the identity of its human operator and incorporate built‑in controls that let users grant or revoke permission for the agent to act on their behalf. While the FTC could not prevent platforms from using non‑compliant providers, it could remove violators from the FTC list.
AI agents—software programs that can make decisions for users—have become increasingly common. A 2025 Morgan Stanley study found that 23 % of Americans used an AI agent to make a purchase within a 30‑day period, and it projected that agent‑driven shopping could reach hundreds of billions of dollars in online commerce by 2030. Yet the technology remains prone to errors: agents can place unintended orders, leak sensitive data, or act contrary to a user’s preferences. The risk is amplified when AI bots interact with one another, creating chains of automated transactions that may be difficult to trace.
“We view this draft as a major step toward building a clear federal framework that promotes innovation, protects consumers, and ensures the United States continues to lead the world in emerging technology,” Warner said. The proposal is open for public comment before a formal version is introduced in the Senate.
The bill arrives amid a broader regulatory push over frontier AI models. Earlier this month the Department of Commerce placed export controls on Anthropic’s Mythos 5 and Fable 5 models. The Trump administration had launched a voluntary 30‑day testing program for frontier models but imposed export controls after Anthropic publicly released Fable 5, citing concerns that the model could be jailbroken. Anthropic has stated that internal testing has not identified universal jailbreaks for Fable 5 and that third‑party research has not shown its guardrails have been circumvented.
While the AI AGENT Act does not address export controls directly, it would establish a domestic framework to ensure that AI agents used on U.S. platforms meet federal security and privacy standards. The FTC’s certification process would be the first step toward a market for trustworthy AI agents, potentially reducing the likelihood of accidental or malicious transactions.
The draft remains a discussion document. Warner’s office has opened the proposal to public comment, and the Senate is expected to consider a formal version later in the year. The FTC has not yet announced specific certification criteria, and no independent bodies have been named. Passage of the bill would signal a significant shift in how AI agents are regulated in the United States and could set a precedent for other countries.
In the meantime, the AI industry continues to grapple with the balance between rapid innovation and consumer protection. The AI AGENT Act represents an effort to codify that balance into a federal framework that could shape the future of AI‑driven commerce and online interaction.
Under the bill, the Federal Trade Commission (FTC) would authorize independent bodies to evaluate AI agent vendors. These bodies would assess whether a product satisfies baseline protections for privacy, data security, and the user’s interests. Each AI agent would also need to link to the identity of its human operator and incorporate built‑in controls that let users grant or revoke permission for the agent to act on their behalf. While the FTC could not prevent platforms from using non‑compliant providers, it could remove violators from the FTC list.
AI agents—software programs that can make decisions for users—have become increasingly common. A 2025 Morgan Stanley study found that 23 % of Americans used an AI agent to make a purchase within a 30‑day period, and it projected that agent‑driven shopping could reach hundreds of billions of dollars in online commerce by 2030. Yet the technology remains prone to errors: agents can place unintended orders, leak sensitive data, or act contrary to a user’s preferences. The risk is amplified when AI bots interact with one another, creating chains of automated transactions that may be difficult to trace.
“We view this draft as a major step toward building a clear federal framework that promotes innovation, protects consumers, and ensures the United States continues to lead the world in emerging technology,” Warner said. The proposal is open for public comment before a formal version is introduced in the Senate.
The bill arrives amid a broader regulatory push over frontier AI models. Earlier this month the Department of Commerce placed export controls on Anthropic’s Mythos 5 and Fable 5 models. The Trump administration had launched a voluntary 30‑day testing program for frontier models but imposed export controls after Anthropic publicly released Fable 5, citing concerns that the model could be jailbroken. Anthropic has stated that internal testing has not identified universal jailbreaks for Fable 5 and that third‑party research has not shown its guardrails have been circumvented.
While the AI AGENT Act does not address export controls directly, it would establish a domestic framework to ensure that AI agents used on U.S. platforms meet federal security and privacy standards. The FTC’s certification process would be the first step toward a market for trustworthy AI agents, potentially reducing the likelihood of accidental or malicious transactions.
The draft remains a discussion document. Warner’s office has opened the proposal to public comment, and the Senate is expected to consider a formal version later in the year. The FTC has not yet announced specific certification criteria, and no independent bodies have been named. Passage of the bill would signal a significant shift in how AI agents are regulated in the United States and could set a precedent for other countries.
In the meantime, the AI industry continues to grapple with the balance between rapid innovation and consumer protection. The AI AGENT Act represents an effort to codify that balance into a federal framework that could shape the future of AI‑driven commerce and online interaction.
