Arctera research finds UK businesses are putting AI governance into practice yet fragmented communications, teams and systems make it difficult to prove how AI-assisted work is safe and compliant.
SAN RAMON, Calif. – November 4, 2026 – UK organisations are making strong progress in bringing AI governance into everyday business processes. However, many still face a major challenge when they need to explain, investigate or defend an AI-assisted decision or communication.
This is according to State of AI Governance 2026: The UK Perspective – Where Operational Governance Meets the Reconstruction Challenge, a new report from Arctera, a business unit of Cloud Software Group and a global provider of compliance and governance solutions.
Moving Beyond AI Policy to Practical Controls
The research shows that UK organisations have put many important AI controls in
place, including human review, employee training, prompt and output logging, and
regular audits.
- 89% require human review or approval for high-risk AI communications, compared with 79% globally
- 85% provide employees with training on compliant AI use versus 77% globally
- 83% have a formal AI usage policy versus 80% globally
- 77% monitor or log AI prompts and outputs compared with 65% globally
- 70% regularly audit AI tools and outputs compared with 63% globally
However, putting controls in place is only part of the story, warns Arctera. Organisations must also be able to produce clear evidence that those controls worked in practice in response to regulator, auditor, legal team or internal investigator questions after an audit, investigation or legal dispute.
‘Reconstruction Challenge’ Prevents Defensible AI Audit Trails
Many businesses face challenges as the information created by these AI controls is often spread across different systems, teams and communications channels.
The report finds that 40% of UK respondents identify correlating communications across different formats and channels as a major challenge.
The risk is particularly high where employees use consumer messaging apps or personal calls. In these cases, 73% of UK organisations say their surveillance coverage is partial or non-existent. The figure is 59% for personal phone calls and 27% for text messages.
As a result, most UK organisations, 80%, are unable to start investigations with all the information they need. And one-third (33%) spend more than half of their investigation time gathering and reconciling information before they can begin analysis.
This creates what Arctera calls the “reconstruction challenge”. When a communication, decision or business record is questioned, there are huge difficulties rebuilding the full sequence of events behind an AI-assisted communication, decision or business record. For many this ends up being a manual, admin-intensive process to piece together AI activity, reviews, approvals, edits and conversations that led to the final outcome.
Soniya Bopache, SVP and General Manager of Arctera
“UK organisations are moving beyond simply setting rules for AI use. They are increasingly putting governance into practice through employee training, human review, prompt and output logging, retention and auditing.
But controls alone are not enough. When scrutiny comes knocking, organisations need to be able to connect the evidence and show quickly and confidently what happened from beginning to end. This is central to defensible AI governance.”
Research Methodology
The UK Perspective is derived from Arctera’s State of AI Governance 2026: The Policy-to-Proof Gap in AI-Assisted Communications, an independent quantitative benchmark conducted by Hanover Research in May 2026 through a pre-vetted industry panel. This UK analysis draws on 100 United Kingdom respondents from the 500-person global study: full-time compliance decision-makers or influencers in finance, healthcare and energy/utilities at organisations with at least 500 employees and US$50 million in annual revenue. Results are unweighted and directional; a traditional margin of error is not applied to nonprobability samples. Full methodology is available in the linked research reports.
About Arctera
Arctera, a business unit of Cloud Software Group, is a leading global provider of compliance and governance solutions that enable firms to unleash game-changing technologies into their organizations while minimizing risk. Created in 2024 from Veritas Technologies, Arctera helps the biggest companies in the world monitor and control exactly how their information is being accessed, used and shared. The ArcteraTM Unified Platform is able to capture data from over 130+ different content sources, and more than 280 AI policies help firms streamline compliance and adapt to evolving regulations.
Arctera and the Arctera logo are either registered trademarks or trademarks ofArctera US LLC and/or its affiliates in the United States and/or other countries. All other product and company names and marks mentioned in this document are the property of their respective owners and are mentioned for identification purposes only.
Learn more at www.arctera.com or connect with us on LinkedIn
Contact: press@arctera.com