When Automation Enters Public Safety Rulemaking

The growing use of artificial intelligence in government decision-making is accelerating,but not without controversy. In early 2026, reports revealed that the U.S. Department of Transportation relied on AI tools to help draft transportation safety regulations. The disclosure immediately triggered backlash from safety experts, legal scholars, and former regulators, who questioned whether algorithm-generated text has any place in shaping rules that directly affect human lives.

At the heart of the debate is a fundamental issue: should AI be involved in writing safety policy at all,and if so, under what safeguards?

What the DOT Actually Did

According to public records and internal documentation, AI systems were used to assist with drafting portions of regulatory language related to transportation safety oversight. Officials characterized the technology as a productivity aid rather than a decision-maker, emphasizing that human staff reviewed the final output.

However, critics argue that even limited AI involvement in rule drafting can introduce hidden risks, especially when the underlying models are trained on opaque datasets and are prone to factual errors, bias, or overly generic reasoning.

Why Safety Experts Are Alarmed

Transportation safety rules are not routine paperwork. They are built on decades of crash data, engineering standards, and legal precedent. Experts warn that AI systems lack:

• Contextual judgment required to weigh trade-offs between economic impact and human safety
• Transparency, since model training data and internal logic are often undisclosed
• Accountability, as responsibility becomes blurred when automated tools contribute to official language. 

Former regulators described the practice as “reckless” and “procedurally unsound,” warning that AI-assisted drafting could weaken enforceability or open regulations to legal challenges.

The Transparency Problem

One of the most significant concerns is disclosure. Critics note that the DOT did not clearly inform the public, at the outset, that AI tools were used in the drafting process. This omission raises questions about compliance with administrative law norms, which rely heavily on transparency, public comment, and traceable authorship.

If stakeholders cannot determine how a rule was written, or whether automated systems influenced its structure, it becomes harder to challenge flawed assumptions during the public review phase.

Legal and Ethical Implications

From a legal standpoint, AI-assisted rulemaking exists in a Gray area. Current U.S. administrative law assumes that regulations are drafted by accountable human officials. Introducing AI complicates that assumption.

Ethically, the issue is even sharper. Safety regulations govern aviation, highways, rail systems, and autonomous vehicles,domains where errors can result in injury or death. Critics argue that delegating any part of this responsibility to probabilistic systems undermines the duty of care expected from public institutions.

Government AI Use: A Broader Pattern

The DOT controversy reflects a larger trend across federal agencies experimenting with AI for drafting memos, analyzing comments, and summarizing technical material. While efficiency gains are real, governance experts stress that speed should never outweigh rigor,especially in safety-critical domains.

Without clear federal standards on acceptable AI use in policymaking, agencies risk inconsistent practices and public distrust.

What Responsible AI Use Could Look Like

Most critics are not calling for a blanket ban on AI in government. Instead, they advocate for strict boundaries, such as:

• Prohibiting AI from drafting binding safety language
• Mandatory disclosure when AI tools are used
• Independent audits of AI systems involved in policy workflows
• Clear human authorship and accountability requirements

These guardrails, experts say, are essential to prevent automation from quietly reshaping public safety governance.

Final Thoughts: Efficiency vs. Responsibility

The DOT’s experiment with AI-assisted rule drafting has become a cautionary example of how emerging technologies can outpace regulatory norms. While AI may help summarize data or organize feedback, its role in shaping safety regulations remains deeply contested.

As governments continue to integrate AI into their operations, the lesson is clear: when public safety is involved, transparency and human judgment are not optional, they are non-negotiable.