Compliance professionals set to increase AI tool spending by 24% – but trust gap keeps humans in the loop, finds S-RM

• 78% expect spending on AI-enabled compliance and due-diligence solutions to increase over the next 12-24 months, with average annual spend forecast to rise 24% to around USD 416,000
• 84% say trust and output reliability factors remain a challenge when implementing AI, highlighting the gap between rising adoption and confidence in AI-generated outputs
• Only 12% expect AI to perform most due-diligence activities with limited human oversight within five years, with 88% expecting human expertise to remain central to the process

London, 5 October 2026: Organisations are increasing investment in artificial intelligence (AI) for compliance and due diligence, but the technology is not yet trusted to replace human judgement in high-risk decisions, according to new research from global corporate intelligence and cyber security consultancy S-RM.

The research, which surveyed 225 professionals with responsibilities spanning compliance and ethics, procurement, supplier management, and third-party and vendor risk, finds that 78% expect spending on AI-enabled compliance and due-diligence solutions to increase over the next 12–24 months. Average annual spend is expected to rise from approximately USD 337,000 today to USD 416,000 in 24 months, a 24% increase.

However, the research points to a significant trust gap. 84% of respondents say trust and output reliability factors remain a challenge when implementing AI, with half (50%) citing concerns about the accuracy of AI outputs and slightly fewer (47%) reporting unpredictable or inconsistent outputs, including hallucinations. Nearly three-fifths (58%) also say AI-enabled due-diligence solutions struggle to assess the context, credibility or relevance of information sufficiently to replicate human-led research.

From screening to due diligence

The findings suggest that the next phase of AI adoption is moving beyond established screening activities and into more complex due-diligence and investigative workflows. Adverse media screening is currently the most common AI use case, used by 48% of respondents, while 38% already use AI for third-party due diligence. Looking ahead, third-party due diligence also tops the list for planned adoption, with a further 48% intending to introduce AI over the next 12 months. This is followed by customer due diligence/Know Your Customer at 42%, enhanced due diligence at 40%, and investigations and case management at 38%.

Organisations are already seeing operational benefits from AI. 81% report a significant or moderate reduction in the time spent on manual research, while 78% report improved analyst productivity and 74% say AI has improved their identification of high-risk entities. Meanwhile, 73% say AI has enabled compliance teams to focus on higher-value work.

The research indicates that organisations are therefore increasingly looking to AI to take on high-volume research and analysis, rather than make consequential decisions independently. Only 12% believe AI will be capable of performing most due-diligence activities with limited human oversight within five years. Instead, 48% expect AI to perform many activities while human experts remain essential for complex decisions, judgement and high-risk cases, while a further 33% expect AI mainly to accelerate research and information gathering as people continue to lead most due-diligence work.

Human oversight is also expected to remain a core part of the compliance operating model, with 72% saying human oversight and assurance will remain essential over the next three years.

AI adoption must keep pace with assurance

For compliance teams, the challenge is therefore shifting from proving that AI can perform individual tasks to integrating it into workflows in a way that is transparent, reliable and defensible. The research suggests that organisations are expanding their use of AI in areas where it can increase the scale and speed of research and analysis, while retaining human expertise to interpret ambiguous information, challenge uncertain outputs and remain accountable for important decisions.

Chris Oehlert, Head of Third-Party Risk Solutions at S-RM, commented:

The market has moved beyond the question of whether compliance teams will use AI. The new question is how they can scale it without compromising decision quality or accountability. The strongest use case is removing repetitive research so experts can spend more time interrogating ambiguous findings, assessing context and making risk-based decisions. As AI moves closer to consequential compliance decisions, transparency, validation and the ability to defend an output become as important as speed and scale.” 

Toby Thomas, Director of Research, Corporate Intelligence at S-RM, commented:

The data shows a clear transition from screening towards more complex due-diligence workflows, where the quality and context of underlying information matter more. AI can process information at scale, but respondents still see difficulties in assessing context, credibility and relevance, which explains why human expertise remains central to the model they expect to emerge.”  

The findings are set out in S-RM’s new report, From Screening to Decision-Making: The Next Phase of AI in Compliance, which examines how organisations are moving from AI-powered screening towards AI-enabled due diligence and decision support. The report also explores how organisations can combine technology, research and human judgement to manage third-party risk at greater scale.

The report is available here: https://www.s-rminform.com/from-screening-to-decision-making-the-next-phase-of-ai-in-compliance