Key Takeaways
- Nearly six in 10 retail leaders say their organizations are adopting AI faster than they can govern it, and only 31% have an AI kill switch.
- Retailers trust AI outputs more than other industries do, 78% vs. 69%, but only 20% completely trust AI to operate without unacceptable risk or to comply with regulations.
Research commissioned by ABBYY shows that nearly six in ten retail leaders say their organizations are adopting AI faster than they can govern it, and just 31% have an AI kill switch to stop a malfunctioning system.
Retailers appear more confident overall in AI than their peers, with 78% saying they trust AI to produce accurate outputs, compared with 69% across other industries. However, that confidence weakens the issues that matter most for governance. Only 20% of retail leaders completely trust AI to operate without introducing unacceptable risks. Similarly, just 20% completely trust AI to comply with applicable regulations, well below the 29% cross-industry average.
The ABBYY Who Answers for AI: The Governance Gap report, conducted by Opinium Research, surveyed 1,200 senior managers in businesses of 100 or more employees located in the U.S., UK, France, Germany, Australia, and Singapore.
What Do Retailers Need to Close the AI Governance Gap?
Roman Kilun, Chief Compliance Officer at ABBYY, noted businesses are embracing AI, but many are still struggling with key questions around accountability, governance and ROI.
“Senior management need to do a better job of communicating what their responsible AI policies are companywide,” said Kilun in a statement. “In addition, AI requires organizations to know where data comes from, whether it can be trusted, how it can be used, and where it is stored and processed.”
Confidence In AI Masks Gaps In Trust, Compliance, Risk
Regulatory uncertainty ranks as the single biggest barrier to achieving greater ROI from AI in retail, cited by 16% of respondents. Integration with existing systems (15%) and the cost of implementation (15%) follow closely, while skills shortages and data quality each rank as one of the top challenges for 12% of retailers. For organizations that handle large volumes of customer, payment, and loyalty data, risks that carry direct consequences for brand trust include
- 43% cite AI-generated cyberattacks as their top AI-related security risk
- 38% are mindful of confidential data leakage
- 32% cite insider misuse of AI.
Human oversight in retail falls short of cross-industry benchmarks. Just over half (51%) of retail leaders say humans review all important AI decisions before action is taken, compared with 56% globally Nearly a quarter (24%) say their AI systems operate largely autonomously, and 14% report no formal approach to human oversight at all, compared with 9% across industries.
Retailers lead on rollback capability, with 45% able to reverse AI changes compared with 35% across industries. Yet only 31% have an AI kill switch to stop a malfunctioning system, below the 35% average.
Retailers Split On Who Answers When AI Goes Wrong
The study reveals that retail leaders do not agree on who should take responsibility when AI systems produce harmful or incorrect outputs. More than a third (of retail leaders at businesses using AI believe the organization and the vendor should share responsibility; 27% think the organization using the AI should bear primary responsibility, while 19% place it with the AI vendor.
More than four in five retail leaders surveyed have an AI governance framework in place, and 39% say their framework is fully implemented and operational, ahead of 30% across industries. Among retailers with a framework, 81% say it has made their AI initiatives more successful, and 58% say it has made scaling AI past the pilot stage easier.
Still, 47% of retailers say their governance framework makes it easier to encourage widespread AI adoption among employees, compared with 59% across industries. In addition, 27% rate their processes for auditing AI systems for compliance, fairness, and transparency as inadequate.
“Organizations we work with are moving beyond pilots and proofs of concept into execution, and clear governance frameworks and a shared understanding of responsibility are critical to building confidence in AI,” stated Kilun.
Access the full 2026 ABBYY Who Answers for AI: The Governance Gap report for more information.





