Published 25 March 2026 · 8 min read

JCQ Head of Centre AI Malpractice Liability Explained

AI misuse in coursework and non-exam assessment is now one of the named concerns of the regulator. In 2025, Ofqual recorded 1,125 cases in which students lost their entire GCSE or A Level qualification through malpractice of any kind (a combined total, not AI-specific). In his 3 March 2026 letter to awarding organisation chief executives, the Chief Regulator named AI misuse in non-exam assessment alongside mobile phones as a key concern — noting that detected AI cases reported to Ofqual remain relatively low but that there is significant worry about the true extent of AI misuse, and calling for stronger institutional controls.

What many Heads of Centre do not realise is that under JCQ regulations, personal liability for AI-related malpractice does not stop with the student. When a school lacks adequate policies, training, or reporting procedures, the Head of Centre can be held directly responsible.

This article sets out the specific JCQ provisions that create that liability, what triggers it, and what schools need to have in place.

The Head of Centre Declaration is personal and non-delegable

Each year, the Head of Centre must sign a formal declaration under JCQ General Regulations, section 1.9. This declaration confirms that the centre adheres to all JCQ regulations, including those governing AI use in assessed work. The deadline is 31 October annually.

This is not a formality. The declaration is personal to the Head of Centre and cannot be delegated to the Exams Officer, a deputy, or any other member of staff. By signing, the Head personally confirms that systems are in place to prevent, detect, and respond to malpractice, including malpractice involving AI tools.

If those systems do not exist when the declaration is signed, the Head has made a false declaration. If malpractice later comes to light, the absence of proper procedures becomes a direct line of liability back to the person who signed.

Failure to report is itself malpractice

Section 1.7 of the JCQ Suspected Malpractice Policies and Procedures 2025-26 makes a point that many school leaders overlook: failure by a centre to notify, investigate and report all allegations of malpractice or suspected malpractice is itself classified as malpractice. This is not negligence. It is not an administrative shortcoming. It is a substantive malpractice finding against the centre and, by extension, the Head of Centre who signed the declaration.

This provision means that a Head of Centre who is aware of concerns about AI-generated work but does not report them, or who presides over a school where staff do not know how or when to escalate such concerns, can face sanctions not just for the underlying student misconduct but for the failure to act on it.

Lack of training is an aggravating factor

Section 6.14 of the JCQ Suspected Malpractice guidance introduces a provision with significant consequences for school leadership. When the Malpractice Committee investigates a case and identifies management failings, such as a lack of staff training on AI detection or the absence of clear internal procedures, it can redirect sanctions from the student to the Head of Centre personally.

In practice, this means that a student caught submitting AI-generated coursework may receive a reduced sanction if the committee concludes that the school failed to train staff adequately, failed to communicate expectations to students, or failed to put procedures in place. The Head of Centre may then face sanctions instead of, or in addition to, the student.

The implication is clear: schools that have not trained staff on AI detection and have not briefed students on what constitutes AI misuse are not just exposed to malpractice findings against individual candidates. They are exposed to findings against the institution and its leader.

The two-phase rule: before and after authentication

Section 4.5 establishes a critical dividing line that determines how AI-related concerns must be handled. The dividing line is the point at which the candidate signs the JCQ authentication statement, the declaration that the work submitted is their own.

Before authentication is signed

If a teacher suspects AI misuse before the candidate has signed the authentication statement, the matter can be dealt with internally. The student can be asked to redo the work, the submission can be rejected, or the teacher can intervene directly. There is no mandatory reporting obligation at this stage.

After authentication is signed

Once the authentication statement has been signed and the work submitted, the situation changes fundamentally. At this point, suspected AI malpractice must be reported immediately via the JCQ Form M1 (Notification of Candidate Malpractice). Internal resolution is no longer an option. The Head of Centre is required to notify the relevant awarding body without delay.

Schools that handle post-authentication AI concerns informally, by quietly asking a student to redo work or by choosing not to escalate, are in breach of JCQ regulations. Under section 1.7, that failure to report is itself malpractice.

The reporting forms: M1, M2, and M3

JCQ uses three forms to structure malpractice reporting. Understanding which form applies in each situation is essential:

Form M1 — Notification of candidate malpractice. Used after the authentication statement has been signed. Reports suspected AI misuse by a student. Must be submitted immediately upon discovery.
Form M2 — Notification of staff or centre malpractice. Used when malpractice involves a member of staff or a systemic centre failing. Unlike M1, there is no pre/post-authentication distinction. M2 must be submitted immediately, regardless of timing.
Form M3 — Investigation report. A follow-up form used to submit the centre's internal investigation findings to the awarding body.

A school that discovers a teacher has been routinely approving AI-generated work without proper checks would need to submit Form M2 immediately. A school that discovers a student submitted AI-written coursework after authentication would submit Form M1, followed by Form M3 after investigation.

All four exam boards defer to JCQ

AQA, OCR, Pearson, and WJEC all operate under JCQ regulations for malpractice (JCQ General Regulations for Approved Centres 2025-26). There is no exam board that applies a lighter standard or offers an alternative procedure. The liability framework described in this article is universal across all awarding bodies used in England, Wales, and Northern Ireland.

This means a school cannot rely on the argument that "our exam board handles it differently." The JCQ malpractice procedures, including the Head of Centre declaration, the authentication dividing line, and the M1/M2/M3 reporting process, apply uniformly.

Sanctions: what is at stake

JCQ Appendix 4 sets out the range of sanctions available to the Malpractice Committee. These apply to both centres and individuals:

For centres

Written warning
Imposition of special conditions on the centre
Suspension of centre recognition (the centre cannot enter candidates for exams)
Withdrawal of centre recognition

For individuals (including Heads of Centre)

Written warning
Requirement to undertake specific training
Suspension from involvement in exam administration
Debarment from exam delivery

Withdrawal of centre recognition or debarment of the Head of Centre are the most severe outcomes. While these are rare, published case examples in Appendix A of JCQ's AI Use in Assessments guidance (Revision 2, April 2025) show that awarding bodies do act. Cases have been published across major exam boards where AI-generated content was identified in submitted work and the candidate was disqualified.

DfE SLT Briefing Pack: the national signal

On 9 March 2026, the DfE's SLT Briefing Pack on AI and Coursework Integrity was published as part of Ofqual's AI in coursework: resources for schools. This document, aimed at senior leadership teams, signals that the DfE expects schools to be actively addressing AI risks in coursework and assessment.

While the briefing pack does not carry regulatory force in the way JCQ regulations do, it reinforces the expectation that school leaders should have considered AI's impact on assessment integrity. A Head of Centre who has neither read the DfE briefing nor implemented any AI-related policy changes would find it difficult to demonstrate adequate due diligence in a malpractice investigation.

What this means for your school

The regulatory framework around AI malpractice is not ambiguous. JCQ regulations create a clear chain of personal accountability from the Head of Centre declaration through to the malpractice reporting process. The key exposures are:

No AI policy in place — The Head of Centre declaration confirms adherence to all JCQ regulations. Without an AI-specific policy, that confirmation is unsupported.
No staff training — Section 6.14 allows the Malpractice Committee to redirect sanctions to the Head when management failings, including lack of training, are identified.
No clear reporting procedure — Staff need to know the two-phase rule (before and after authentication) and which form to use. Without this, the school risks a failure-to-report finding under section 1.7.
No student briefing — Students must understand what constitutes AI misuse before they sign the authentication statement. Without clear communication, the school shares responsibility for any resulting malpractice.

Schools that have addressed these four areas, with documented policies, recorded training, and clear procedures, are in a materially stronger position if a malpractice case arises. Those that have not are carrying a level of institutional and personal risk that increases with every exam series.

Next steps

If you are not sure whether your school has adequate AI policy coverage, start with a practical self-assessment. Our free AI Policy Checklist for School Leaders covers the nine areas that matter most, each item referenced to current published guidance from DfE, JCQ, and ICO.

If you already know there are gaps, the AI Policy Starter Pack provides source-referenced policy documents, including a Staff Acceptable Use Policy, Student Acceptable Use Policy, and Assessment Integrity Policy, designed to address the JCQ requirements outlined in this article. All documents reflect current published guidance as of March 2026 and are provided for review and adaptation by the school or trust.