AI is present in reporting, but humans still lead
Corporate reporting remains steadfastly human-led even as artificial intelligence (AI) adoption gathers pace, according to new research published by the UK’s Financial Reporting Council (FRC).
The study, conducted by Lancaster University on the FRC’s behalf, draws on interviews and a survey of preparers, investor relations teams, advisors and auditors. It finds that generative AI is seen as a valuable tool in many teams, with 39% of organisations reporting current use and a further 31% piloting AI. However, deployment is concentrated in lower-risk, task-specific activities such as drafting, data extraction, reconciliation, anomaly detection and consistency checking. High-judgement areas, including management commentary and forward-looking statements, remain squarely in the human domain, with interview participants stressing that AI is used for first-draft work, supervised by human eyes.
Why the caution? Preparers point to factors ranging from trust and reputational risk to unreliable underlying data. The study highlights that in many organisations the data drawn upon for reporting “remains fragmented or inconsistent,” putting the brakes on more extensive AI use. For now at least, AI is shaping up to be a valuable assistant to a human-led reporting pipeline.
High-quality, validated, machine-readable reporting could help provide the reliable data foundation that AI tools need, whether extracting, analysing or drafting. Questions also remain to be answered around audit and responsibility as AI use in corporate reporting grows.
The research forms part of the FRC’s efforts to understand how technology is reshaping the reporting landscape so that regulation stays responsive and proportionate, alongside its ongoing Audit Tech and AI Sandbox.
Read the research here.
