On March 18, 2026, U.S. Senator Marsha Blackburn issued an artificial intelligence (“AI”) legislative framework discussion draft (“AI Discussion Draft”) titled the “Trump America AI Act.” According to Senator Blackburn, this intends to codify President Trump’s December 11, 2025, executive order for establishing a uniform federal AI policy. Senator Blackburn stated, “[President Trump] called on Congress to pass federal standards and protections to solve the patchwork of state laws that has hindered AI innovation.” The AI Discussion Draft, per Senator Blackburn, is intended to broadly protect “children, creators, conservatives, and communities from exploitation, abuse, and censorship and ensure American AI companies can innovate without cumbersome regulation.”
Scope
The AI Discussion Draft, as written, establishes a wide‑ranging federal framework for AI, organized into 17 titles that fall into several major thematic areas. As summarized by Sen. Blackburn’s office, the AI Discussion Draft includes:
- Safety and Protection measures include developer duties of care for AI chatbots, safeguards for minors through the Kids Online Safety Act, restrictions on AI companions for minors, and requirements for transparency, provenance, and watermarking of synthetic content.
- Governance, Accountability, and Liability provisions create reporting obligations for AI‑driven job impacts, repeal Section 230 of the Communications Act, define liability standards for developers and deployers, require audits of high‑risk AI systems, and mandate registration of foreign AI developers.
- Risk Management and National Security components establish a federal evaluation program for advanced AI systems, and address risks such as loss‑of‑control scenarios, weaponization, threat to critical infrastructure, and erosion of civil liberties.
- Innovation, Research, and Standards aims to promote voluntary AI standards, testbeds, and international cooperation, establish the National AI Research Resource, expand public data resources, and implement federal grand challenge prize competitions.
- Intellectual Property, Copyright, and Digital Identity reforms include voice and likeness protections, subpoena authority for AI training data usage, and provisions seeking to establish that AI training and certain derivative uses are not fair use under the Copyright Act.
- Federal Agency AI Use and Infrastructure provisions require agencies to procure “unbiased” AI models and impose protections for electricity ratepayers regarding data‑center infrastructure owned by private companies.
Next Steps and Takeaway
As a discussion draft, it is unclear whether Sen. Blackburn’s bill will begin standard legislative proceedings, such as presentation to appropriate Senate committees, or whether – true to its name – the next steps are more informal stakeholder discussions. If the latter, the next steps would generally involve gathering feedback from federal agencies and private sector organizations that may be affected by the proposed legislation, in addition to engaging with other Senators in efforts to build a viable coalition of sponsors and support. Given the breadth and complexity of the current AI Discussion Draft, stakeholders across the political spectrum and the technology ecosystem should anticipate an extended period of negotiation, amendment, and potential restructuring before any final legislative text emerges.
For more information on AI legislation, regulations, and enforcement, please contact Alston & Bird’s Privacy, Cybersecurity and Data Strategy Team and sign up for alerts at AlstonPrivacy.com.
