AI Regulation Tracker  /  Research and insight

FRC Research Finds AI in UK Corporate Reporting Is Rising but Still Human-Led

On July 8, 2026, the UK's Financial Reporting Council published research it commissioned from Lancaster University finding that AI use in corporate reporting is rising but cautious and uneven, with reporting still human-led. This is research and insight, not a rule, and it changes no legal duty.

By Anthony Guerriero  ·  The Leveraged Years Regulatory Research  ·  Published July 15, 2026  ·  Last updated July 15, 2026

On July 8, 2026, the Financial Reporting Council published the results of research it commissioned to understand how corporate reporting is adapting to AI. The work was carried out by a team at Lancaster University, with academics from Loughborough University. It draws on 39 interviews with senior people in reporting and investor-relations functions, along with advisers, software vendors, and auditors, plus an online survey that returned 103 responses. The survey was open from February 17 to April 21, 2026.

What the research actually found

The top-line finding is straightforward. AI use in corporate reporting is going up, but the adoption is cautious and uneven across the market. There is no wave of companies handing the annual report to a machine. What the research describes instead is measured, targeted use in specific spots.

Where AI shows up most is the narrative side of reporting. Generative AI is emerging as a useful tool for drafting narrative content, and the research frames that use as first-draft work rather than finished analysis. Think of the tasks that are repetitive and grounded in existing numbers: roll-forwards, describing what a chart or a graph shows, compliance-driven passages that follow a known pattern year to year. That is where the tools are earning their place. The higher-value, opinion-based commentary, the parts of the report where a company is actually making a judgment call, stays with people.

Use inside the financial statements themselves remains limited. That is the part of the report where the numbers have to be right and the professional judgment is heaviest, and it is the part where companies are moving slowest. The research characterizes deployment across the board as concentrated on lower-risk, task-specific activities rather than areas that require significant professional judgment.

The workflow the research describes is human-led. Generative AI is being used as an assistant inside a review-and-edit process run by people, not as an autonomous author that produces final copy on its own. A person still owns the output, still checks it, and still signs off on it.

The research also captured why companies are holding back. Preparers pointed to trust, data quality, governance controls, and concerns about legal, reputational, and stakeholder risk as the main barriers to using AI more widely. Those are not abstract worries. They are the same questions any finance leader would ask before letting a model near a number that investors and regulators are going to read.

What this is, and what it is not

I want to be precise here, because it is easy to over-read a regulator's name on a press release.

This is research. The FRC commissioned it to get a picture of where the market is, and it reads it as a snapshot of the current stage of adoption. It is not a new rule. It is not a change to a reporting standard. It does not impose a disclosure obligation, a control requirement, or a filing duty on anyone. Nothing in a preparer's or an auditor's legal responsibilities changed on July 8 because this report came out.

What the research does is set expectations and inform thinking. When the body that oversees corporate reporting and audit in the UK publishes a considered view that AI is being used cautiously and that reporting remains human-led, that tells you how the regulator currently sees the landscape. It is a weathervane, useful for calibrating your own approach and for anticipating where guidance or scrutiny might eventually go. It is not a wind you are legally required to turn into.

That distinction matters most for anyone tempted to point at this report and say the FRC has "approved" or "endorsed" a particular way of using AI in reporting. It has not. It has described what companies are doing and flagged the risks those companies themselves raised.

What this means for auditors and finance teams

For UK reporting teams, the research is close to a mirror. If you are using generative AI to speed up narrative first drafts while keeping people firmly in control of judgment and of the financial statements, you are doing roughly what the market is doing and what the regulator has just described as the prevailing pattern. If you are further ahead than that, the report is a reminder that the barriers your peers cited, trust, data quality, governance, and legal and reputational risk, are the ones a regulator is now paying attention to.

For US CPAs and finance leaders, the value is comparative. The FRC is not your regulator, and this report carries no weight over US filings. But it is a clear, current read from a serious English-speaking authority on where AI is actually landing in the reporting process, and the pattern it describes travels well. AI as a drafting aid for repetitive narrative content, humans in charge of judgment, and heavy caution around the numbers is a posture that lines up with how US standard-setters and audit regulators have been talking about the same tools. Treat it as a data point that helps you benchmark your own controls, not as anything that binds you.

The practical thread running through all of it is governance. The barriers preparers named are the same ones that turn into audit findings and restatements when they are ignored: where did the data come from, who checked the output, and can you show your work. The research does not tell you to build those controls. It tells you that the people already using AI in reporting see them as the gating issue.

What to do now

Read the actual finding, not the headline. The takeaway is that AI use is rising but cautious and still human-led, not that a regulator has cleared AI for corporate reporting or restricted it. Map where AI already touches your reporting, since narrative drafting, chart descriptions, roll-forwards, and compliance boilerplate are where the research found most of the activity. Keep a named person accountable for every AI-assisted passage, especially anything involving judgment or anything inside the financial statements. Write down your governance answers before you need them: data provenance, review steps, and who signs off. And do not treat this report as a rule. If you change a control or a disclosure, base that on your own applicable standards and your auditor's advice, not on a UK research paper.

Browse the full AI Regulation News tracker

Informational analysis for working professionals, not legal, accounting, or audit advice. Confirm how any standard or requirement applies to your situation with qualified professionals in the relevant jurisdiction.