A Draft CCAB Statement Says Professional Accountants Cannot Abdicate Responsibility for Outputs Produced by Their Use of AI

The Consultative Committee of Accountancy Bodies (CCAB) is registered in England and Wales, and its draft lists ICAEW, ACCA, CIPFA, ICAS and Chartered Accountants Ireland under its name. In June 2026 it published a draft statement to the accountancy profession on the ethical use of artificial intelligence, marked as draft guidance "for consultation." The statement "applies to all professional accountants, whether in public practice or business," and covers "the entire AI lifecycle from development and procurement to deployment and monitoring." It is a draft, and it may change. Parts of it follow, not the whole.

A Statement Prepared With AI

The draft begins with a foreword and disclosure. In the interests of transparency and professional integrity, the committee "discloses that AI tools were used, alongside human expertise and professional judgement, in the preparation of this Statement." Responsibility for "the content, conclusions and judgements expressed rests entirely with CCAB."

Its questions and answers say how. AI tools were instructed to use the CCAB Statement on the Use of Sanctions as a template, and were given links to the international code of ethics for accountants and to resources produced by professional accountancy bodies, including IFAC and the committee's members. They were also given an initial draft by the CCAB Ethics Group. A composite of their drafts was then "refined and quality assured" by members of the ethics group, and the statement was quality assured by the group and by technical specialists in each member body before publication. The ethics group, the draft says, considered that using AI to write a statement about the ethical use of AI "would generate engagement and discussion."

Measured against its own advice on transparency, the disclosure goes part of the way. It says who checked the draft and who answers for it. It does not name the tools or say how much of the text came from them.

Risks and Responsibility

The draft says the use of AI tools "may create, exacerbate or obscure ethical risks." Among the risks it says are commonly identified are automation bias, "a tendency to favour automated outputs over human reasoning, even when such output is contradictory or unreliable," and the risk of inputting confidential data, including commercially sensitive client data, into public AI models. AI tools, it says, "may produce factually incorrect but plausible and authoritative sounding outputs."

The potential for the risks it lists to crystallize increases, it says, where AI tools are used "at scale, at speed, or with limited human oversight." It reminds professional accountants "that they cannot abdicate responsibility for outputs produced by their use of AI tools and systems."

What It Asks of Accountants

The draft applies the code's fundamental principles to AI use. Under integrity, accountants should not recklessly associate themselves with AI outputs without adequate validation. Objectivity, it says, "requires scrupulous checking of AI-quoted sources to mitigate automation bias." Under confidentiality, client or any sensitive data "should never be input into an AI tool without specific consent," particularly where the tool may retain data for training, or where its use might put the data into the public domain.

Under professional behavior, compliance with that principle, the draft says, "also includes ensuring that AI generated outputs have been fully checked and validated before submitting them to a Court or placing them in the public domain," or publishing them where others may rely on them.

Accountants are required, the draft says, to apply the code's conceptual framework "to identify, evaluate and address threats to compliance with the fundamental principles arising from AI use." And: "Where threats cannot be eliminated or reduced to an acceptable level through safeguards, the professional activity must be declined or discontinued." In its section on the public interest, the draft says accountants should advocate for AI systems that are fair, accountable and transparent, which involves challenging vendors to provide model cards that specify risks and biases, and ensuring AI use does not propagate disinformation or fraud. On AI agents, it says that "While tasks can be delegated to an agent, the professional accountant remains professionally responsible for the outcome obtained by the use of an agent."

Its practical tips include "Never take an AI output at face value." Another tells accountants to be transparent with clients about how AI is used on the engagement, "and state this clearly in the engagement letter."

Professional accountants are encouraged, the draft says, to comment "to improve the quality of the document before it is finalised."