Compliance teams risk fines when skipping AI checks

Enhesa has found that compliance teams that skip expert checks on AI-generated regulatory guidance are far more likely to face fines. Its survey also found that 56% of teams reported some form of negative outcome linked to unverified or incorrect AI output.

The research points to broad use of AI across multinational compliance functions as companies struggle to keep pace with regulatory change. Enhesa tracked more than 22,000 major regulatory changes in 2025 and more than 15,000 in 2026 through August.

A survey of 273 senior compliance, risk, environment, health and safety, legal, and sustainability decision-makers at companies with 5,000 or more employees found that 40% of compliance teams always or often act on AI-generated regulatory guidance without expert review. Teams that skipped verification entirely were nearly three times as likely to report a regulatory fine, at 37%, compared with 13% for teams that did not.

Separate findings showed that teams which rarely or never skipped expert review reported a regulatory fine rate of 7%, compared with 27% for all other teams. The results suggest that AI's speed gains are being offset by higher operational and regulatory risk when organisations reduce human oversight.

Growing pressure

The backdrop is rising regulatory complexity across large international businesses. Nine in 10 compliance teams said they had experienced at least one business impact from regulatory complexity in the past year, including higher legal or external consulting costs, heavier workloads, and delayed product launches or market entry.

That pressure appears to be influencing expansion plans. Sixty per cent of companies said growing complexity had led them to reconsider, delay, or abandon entry into a new market in the past two years, while 18% said they had abandoned a market entry outright.

Enhesa found that 88% of respondents said regulatory requirements changed again before the previous change was fully implemented. Against that backdrop, 76% said they now use AI across the entire compliance function or run it fully autonomously on at least one core compliance task.

Respondents said AI has delivered most clearly on speed rather than foresight. Forty-eight per cent reported significant improvement in the speed of identifying and responding to regulatory changes, and the same share cited gains in the efficiency of compliance monitoring and reporting. Only 26% reported similar improvement in anticipating future regulatory risk.

Shadow AI

The survey also highlighted concerns about the use of personal AI tools for sensitive work. Sixty-four per cent of compliance teams said they had entered sensitive material into personal or non-corporate AI tools in the past year, most often proprietary chemical or manufacturing data.

Those teams reported a much higher rate of negative outcomes. Eighty-two per cent of respondents who entered sensitive material into personal or non-corporate tools said they had experienced a negative outcome, compared with 9% of teams that entered nothing.

Across all respondents, the most commonly cited consequences of unverified or incorrect AI output were incorrect filings, reported by 30%; regulatory fines, at 21%; missed deadlines, at 20%; and failed audits, at 18%.

The risks were higher among teams that had moved furthest towards autonomous AI. Of the 37% running fully autonomous or agentic AI on at least one task, 71% reported a negative outcome, compared with 47% of all other teams.

Need for audit trails

The findings suggest that many compliance leaders know what evidence they need from AI systems but do not yet feel equipped to produce it consistently. Ninety-one per cent said source citation was critical before acting on AI output, while 88% wanted a full audit trail showing how the system reached its conclusion.

Even so, only 32% said they were very confident that their team could defend the logic behind an AI-driven compliance decision. That gap may prove important as businesses are asked to explain not only the outcome of a compliance decision, but also the process behind it.

The report also sheds light on how leaders want to balance automation with human judgement. Forty-three per cent said AI paired with expert human counsel would be the most effective model for helping their team stay compliant over the next 12 months, ahead of AI tools alone at 39% and expert human counsel alone at 4%.

Keith Berry, chief executive officer of Enhesa, said rising regulatory churn was already creating direct business costs. "Regulatory change is moving faster than anyone can track manually and comes with material business impact. Fifty-three per cent of companies said they increased outside consulting costs, and 34% delayed a product launch or market entry as a result," he said.

He added that the value of AI depends on the quality and defensibility of the information behind it. "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," he said.

On how companies may need to structure their compliance work, Berry said: "AI alone provides teams with speed. Expert counsel alone provides strategic, defensible judgement. Our data shows that together, compliance teams can move as fast as regulatory change demands, with the confidence that every decision will hold up."