AI CPD · barristers · the Established Practitioners Programme

AI competence is a CPD objective. Record it properly.

Barristers’ CPD is self-directed: you set the objectives, you choose the learning, you certify the record. Here is how AI training fits that structure, what it should cover, and why the case law has made it a defensible objective for this year’s plan.

How barristers’ CPD actually works

Since the Bar Standards Board replaced accredited hours with the Established Practitioners Programme, CPD at the Bar has been self-directed. A barrister reflects on their practice, sets written learning objectives for the year, completes activities that serve those objectives, and keeps a record they can produce if the BSB asks for it. There are no mandated courses, no points, and no accreditation of providers.

That structure cuts both ways. It gives members complete freedom over what to learn, and it puts the burden of choosing well, and recording honestly, on the individual. An objective is only as good as the record behind it.

Why AI belongs in this year’s objectives

Three things changed the calculation. First, the failures reached the courts: in Ayinde v Haringey [2025] EWHC 1383 (Admin) the Divisional Court dealt with fabricated authorities in filed documents, and its judgment made clear that those with leadership responsibilities, including heads of chambers, are expected to take practical and effective measures on the safe use of AI. Second, the regulator moved: the BSB’s Guidance on the use of AI and Other Technologies came into force in May 2026, reading AI into the existing Core Duties, from competence to confidentiality to the duty to the court, and the Bar Council maintains its own generative-AI guidance alongside it. Third, the tools stopped being avoidable: instructing solicitors use them, opponents use them, and members of every set already use them privately, trained or not.

Against that background, “understand generative AI well enough to use it safely in practice” is about as defensible as a learning objective gets. It is specific, it is current, and the regulator’s own guidance points at it.

What good AI training for barristers covers

A session worth recording against a CPD objective covers the ground a practitioner actually needs:

  • Capabilities and failure modes. What these systems do well, and precisely how they fail, because invented citations and confident wrongness are the failures that end up before a judge.
  • The guidance, plainly read. What the Bar Council guidance and the BSB’s expectations amount to in practice, without either alarmism or salesmanship.
  • A verification workflow. A repeatable habit for checking AI-assisted research and drafting before it goes anywhere that matters. This is the single highest-value hour in any programme.
  • Confidentiality and privilege boundaries. What must never go into a consumer AI account, and what a chambers-safe setup looks like instead.
  • Hands-on practice. Realistic exercises from the members’ own practice areas, because judgement about the tools only comes from using them under guidance.

Recording it

For the CPD record, a structured session gives you the pieces the EPP asks for: the objective it served, the provider and format, the learning outcomes, and your own reflection on what changes in practice. We provide a session outline and written learning objectives with every workshop for exactly that purpose. To be clear about the boundary: no AI course at the Bar is “accredited”, ours included, because no accreditation exists. What a good course is, is structured to support the CPD framework barristers already work within.

Neura North runs live, on-site AI workshops for barristers’ chambers: a £750 leadership briefing and a £2,500 half-day for the whole set. The clerks’ room side of the story, automating chambers administration itself, lives on our barristers’ chambers page.

01 / Asked before booking

AI and CPD, answered straight.

Does AI training count as CPD for barristers?

It can. Under the Bar Standards Board’s Established Practitioners Programme, barristers plan their own CPD against learning objectives they set themselves, then record and self-certify the activity. There is no list of approved courses and no points system, so AI training counts where a barrister identifies AI competence as a learning objective, undertakes relevant training and records what it achieved. Structured sessions with clear objectives make that record straightforward to keep.

Is there a formally accredited AI course for the Bar?

No accreditation scheme for CPD courses exists at the Bar, because the Established Practitioners Programme deliberately moved away from accredited hours in favour of self-directed learning. Any provider claiming a course is “BSB accredited” is misdescribing how barristers’ CPD works. The honest description of any AI course, including ours, is that it is structured to support CPD requirements: clear objectives, substantive content and a record members can certify against.

What should AI CPD for a barrister actually cover?

Four things, at minimum: what generative AI does and where it fails, because the failure modes, invented authorities above all, are what reach the courts; what the Bar Council and BSB guidance asks of practitioners; a working verification habit for any AI-assisted research or drafting; and the confidentiality boundaries, since client material and legal professional privilege do not belong in consumer AI accounts. Hands-on practice matters more than slides, because judgement about the tools only develops by using them.

Why is AI competence a sensible CPD objective now?

Because the courts have already dealt with what happens without it. In Ayinde v Haringey [2025] EWHC 1383 (Admin), the Divisional Court confronted fabricated authorities in filed documents and made clear that leadership, including heads of chambers, is expected to take practical and effective measures on safe AI use. Members are using these tools whether or not chambers has a position, and a recorded, structured piece of CPD is the demonstrable response.

Can a whole chambers do AI training together?

Yes, and it tends to work better that way, because chambers-wide sessions surface the questions individual study misses: what the set’s own policy should say, how clerks and members divide responsibility, and which workflows AI should never touch. A live session in chambers also lets members work on realistic scenarios from their own practice areas rather than generic examples.