AI adoption in compliance has become nearly universal as regulatory complexity rises. While the technology offers many benefits, it must be reviewed for errors. However, many companies seem to be skipping this step.

A new survey, The state of AI in regulatory compliance: 2026 benchmark report, from Enhesa found that 40% of compliance teams act on AI-generated regulatory guidance always or often without expert review.

The consequences of this could prove costly since teams that skip verification entirely are nearly three times as likely to report a regulatory fine (37%) against those that don’t (13%).  

Enhesa tracked more than 22,000 major regulatory changes in 2025, along with over 15,000 major changes in 2026 through August, and the study is based on responses from 273 senior compliance, risk, EHS, legal and sustainability decision makers at multinational companies with 5,000 or more employees.

To manage regulatory pressure, 76% of teams now run AI across the entire compliance function or fully autonomously on at least one core compliance task. Fifty-six percent (56%) have already faced a negative outcome tied to unverified or incorrect AI output.

"Regulatory change is moving faster than anyone can track manually and comes with material business impact," said Keith Berry, CEO of Enhesa. "Fifty-three percent of companies said they increased outside consulting costs, and 34% delayed a product launch or market entry as a result. AI can help companies stay ahead of regulatory change, enter new markets faster and protect their bottom line, but only when leaders can trust the data informing it and defend their actions.”  

Additional findings from the report include:

Regulatory complexity is straining compliance teams

Ninety percent (90%) of compliance teams report at least one business impact from regulatory complexity in the past year, such as increased legal or outside consulting costs (53%), a significant increase in workload (41%) and delayed product launches or market entry (34%).  

Growing complexity has caused 60% of companies to reconsider, delay or abandon entry into a new market in the past two years, including 18% who abandoned one outright.

Eighty-eight percent (88%) say regulatory requirements change again before the previous change is fully implemented.  

AI delivers speed more than foresight

Compliance leaders report significant improvement over the past 12 months from AI use in speed to identify and react to regulatory changes, as well as in efficiency of compliance monitoring and reporting – each cited by 48% of respondents.

Improvement in speed from AI use is reported nearly twice as often as improvement in anticipating future regulatory risk (26%).

Unverified AI outputs harm business

Negative outcomes reported in the past 12 months attributed to unverified or incorrect AI output include incorrect filings (30%), regulatory fines (21%), missed deadlines (20%) and failed audits (18%).

The gap widens for teams furthest along with AI. Of the 37% running fully autonomous or agentic AI on at least one task, 71% reported a negative outcome, compared with 47% of everyone else.

Shadow AI use is linked to a higher rate of negative outcomes

Sixty-four percent (64%) of compliance teams put sensitive material into personal or non-corporate AI tools in the past year, most often proprietary chemical or manufacturing data (33%).

Those same teams also reported negative outcomes at 82%, compared with 9% of teams that entered nothing.

Defensibility standards exist and are known, but go unmet

Ninety-one percent (91%) of leaders call source citation critical before acting on AI output. Eighty-eight percent (88%) want a full audit trail showing how the AI reached its conclusion.

Just 32% of compliance leaders are very confident their team can defend the logic behind an AI-driven compliance decision.

Teams that verify AI outputs avoid fines

Teams that rarely or never skip expert review reported a regulatory fine at just 7%, compared with 27% of everyone else.

Forty percent (40%) of compliance teams act on AI-generated regulatory guidance always or often without an expert reviewing it first. The top reasons are pressure to improve speed and efficiency (47%), confidence in the AI tool’s accuracy (42%) and regulatory deadline urgency (34%).

Forty-three percent (43%) of leaders say AI paired with expert human counsel would be most effective in helping their team stay compliant over the next 12 months, more than AI tools alone (39%) and 10 times the share who chose expert human counsel alone (4%). Overall, 82% of leaders chose an option involving AI.