ABA’s Guiding AI Principles shape our advocacy on behalf of members. Our long-term vision is for AI to be held in public trust and accountable to the people it affects. For now ABA is focused on these more immediate actions.

Most Immediate: Oppose the SANDBOX Act 
(Currently in committee)

ABA strongly opposes the SANDBOX Act. We join many others who are concerned that the bill’s broad definitions would effectively shield AI companies from future regulations and consumer protections. It was introduced by Senator Ted Cruz in September 2025 and referred to the Senate Commerce Committee where it has not progressed. However, it was named in the White House’ March 2026 AI legislative map and since Cruz chairs the committee it sits in, he can control what happens next. The Act allows AI companies to apply for waivers from federal rules that interfere with their “progress”. Waivers can last up to 10 years, including renewals, and gives the White House, not experts, control over appeals related to potential harm and then, harm is defined too narrowly. 

Other Priorities
Note: Although much of the below legislation is currently stalled, we expect new similar legislation to be introduced in 2027. 

Support Transparency

Support the Transparency and Responsibility for AI Networks Act (TRAIN Act) (Currently sitting in committee)
ABA supports this act which would give authors and other creators a way to find out whether their work was used to train AI without their consent or compensation.

Support AI Foundation Model Transparency Act of 2026 
ABA supports this act that would require the Federal Trade Commission (FTC) to disclose the use of data for training.  

Support Protecting Consumers from Deceptive AI Act 
ABA supports this act that establishes standards for labeling AI-generated content. 

Support the CLEAR Act 
(Currently sitting in committee)
ABA supports the Copyright Labeling and Ethical AI Reporting Act (CLEAR Act), which would require AI developers to notify the US Copyright Office of the copyrighted works used to train their models. Those notices would go into a public database, and creators could take developers to court if they don’t comply.

Support Federal Training Data Transparency Requirements
ABA supports federal requirements for AI developers to publicly disclose the types of data used to train their systems, modeled on California’s AB 2013. A national standard would protect creators in all 50 states and US territories.

Support Opt-In 

Support Licensing Frameworks 
ABA supports opt-in licensing frameworks at the federal and state levels. These would require AI developers to get permission from creators and pay them before using copyrighted works in training data.

Support COPIED Act
ABA supports this act and other content provenance standards and limits on training with protected content without consent, including a certified symbol or mark showing when an AI model was trained on licensed material. The certification could come from a private company or the US government, but it must be regulated and universally adopted.

Support Data Center Regulation

Support AI Data Center Moratorium Act
ABA supports this act that would pause new or expanded AI data centers until certain federal laws pass.

Support Overall Regulation
ABA supports regulation to ensure transparency around the building of data centers and to ensure that they are built responsibly and that communities and the environment are protected. Legislation should include:

  • Nondisclosure agreements between governments and data center developers that hide key information, such as ownership, resource demands, environmental impact, and public subsidies or incentives.
  • A requirement of comprehensive data center impact reports and the empowerment of local authorities to deny projects and/or require mitigation based on the findings of those reports.
  • A requirement that data centers be self-sustaining and pay the full cost of the infrastructure and resources required to build and run the facility.
  • A requirement that developers set aside funds, through escrow or bonds, to cover demolition and site cleanup if a data center closes.
  • A regulation that reduces the environmental impact of AI data centers.
  • Federal incentives that push AI companies to find better solutions.
  • Federal incentives for alternative smaller, distributed data centers.
  • Federally funded research to address environmental problems caused by data centers.

Join Us

Members can support ABA’s advocacy by:

  • Signing up for our VoterVoice alerts, to make it easy to write your representatives about important legislation.
  • Voting for candidates who support AI regulation as outlined above
  • Supporting representatives who support AI regulation.
  • Attending city council meetings to stay informed about proposed local data centers and to call for transparency and accountability.
  • Using bookstores as an education resource for customers: creating displays and hosting events on this topic.