Civil society coalition urges US Congress to reject SANDBOX Act exemptions for AI companies
More than 80 organisations have asked US lawmakers to oppose Senator Ted Cruz’s SANDBOX Act, warning that the proposal would allow artificial intelligence (AI) developers to secure long-term exemptions from federal rules, weakening consumer, civil rights, labour and environmental protections.
A broad coalition of civil society organisations has called on the US Congress to reject the Strengthening Artificial Intelligence Normalization and Diffusion by Oversight and eXperimentation Act, better known as the SANDBOX Act. The proposal, introduced by Senator Ted Cruz, would enable companies developing artificial intelligence systems to request exemptions from existing federal regulations for up to ten years.
In a joint letter submitted on 9 December 2025, the Coalition for Sensible Safeguards and more than 80 allied organisations, including the Center for Democracy and Technology and the Electronic Privacy Information Center, warned that the bill prioritises deregulation over public protections. According to the submission, the legislation would allow developers to bypass key safeguards by granting the Office of Science and Technology Policy authority to issue regulatory waivers in two-year increments.
The organisations argue that the bill’s scope is overly broad, potentially covering areas such as civil rights, labour standards, consumer protection, and environmental rules. They note that the proposal places applicants in the position of leading risk assessments, limiting the role of regulators and excluding those most affected by potential harms. The letter also raises concerns about limited obligations for companies participating in the programme, which would only be required to issue general public notices that their AI systems may pose risks.
The coalition emphasises that the SANDBOX Act could create a pathway for temporary exemptions to become permanent regulatory rollbacks. The bill directs the Office of Science and Technology Policy to recommend that Congress amend or repeal regulations waived under the programme, supported by fast-track legislative procedures.
Civil society groups argue that such an approach would erode accountability mechanisms for AI systems at a time when stronger oversight is needed. They caution that marginalised and vulnerable communities could be disproportionately affected by deregulated AI deployments, particularly in areas such as hiring, lending, public services and environmental exposure.
The coalition’s appeal urges lawmakers to reinforce, rather than weaken, federal oversight frameworks and to ensure that emerging AI technologies are governed through standards designed to protect public welfare and prevent discriminatory or unsafe outcomes.
