AI regulatory requirements for the Bar.

Why the BSB acted now

As Orlagh explained at the outset: "It's useful probably to point out that this is the first time the Bar has had explicit and specific guidance from the regulator around how to manage the very fast-paced and evolving AI technology that is coming at us from every direction."

The backdrop is clear. Ewen noted that research published last year found: "People were interested in new technology, but that there were sometimes challenges in terms of knowing what to do or knowing how best to adopt a new technology. And often what people were doing was kind of sitting back to see how it beds in and see how other people were using it."

Ewen stated plainly: "Judges are commenting in court about the need for more guidance on AI. And we are starting to get reports about barristers as well."

What courts can do: the Ayinde precedent

If you haven't read Ayinde, it's worth a read. Orlagh explained its significance: "The Ayinde case is very, very interesting in terms of what it says about the judiciary's expectations of barristers and solicitors. But barristers feature quite heavily in this case, and it's useful to look at. One of the things that was interesting to me was that so vexed were the judiciary in this case, it would feel when you read it that they actually pointed out all of the powers that the court had against lawyers as such. And this is what they said that they could do: Public admonition, cost orders, wasted cost orders, striking out a case, referral to a regulator, contempt proceedings and in egregious situations, referral to the police."

The core duties remain unchanged — but apply to AI

There are no new rules. Ewen explained: "The real focus for us is on the existing rules and core duties that exist in the handbook...at the end of the day, if you use a technology, ultimately you are responsible for the output of that technology. So, you need to bear the same ethical considerations in mind that you would have if you weren't using that kind of new technology."

Orlagh reinforced this: "The core duties actually remain unchanged. AI doesn't alter the fundamental professional obligations. And therefore, one of the questions that we had from someone was, well, when's the start date for this? And I think the point is there's no start date because there's nothing new. You have to just make sure that in the world of AI, which does exist, whether we like it or not, you are meeting your core duties currently."

Competency is the priority

The AI Guidance from the BSB is signalling that competency in AI is non-negotiable.

The guidance creates a competency requirement that applies across the profession. Ewen stated: "Essentially what we're talking about is making sure that if you're using new technologies that you're competent to do it, you understand it, and you're not generating new risks for clients by doing so."

This applies even if you've chosen not to use AI tools yourself. The profession operates as an ecosystem. Your instructing solicitor will use AI. Your client will use it. Opposing counsel will use it. Litigants in person will use it. Orlagh pressed on this reality: "Given the ecosystem that you all work within in terms of any case, it doesn't take a lot of thought to suddenly realise your instructing solicitor might use it, your client might use it, opposing counsel might use it. If you're against a litigant and person, there's a very good chance that they might use it. Is that the basis, I guess it seems to be from the guidance that in an era of AI technology, it's a fundamental professional obligation to understand it."

The competency requirement is about literacy. Being able to spot when AI has been used. Understanding what it can and cannot do. Recognising hallucinated citations. That's what the profession needs from every barrister, regardless of whether they use AI tools in their own practice.

Documentation and evidence matter in enforcement

When the BSB investigates misconduct, Ewen explained the approach: "If we receive a report about misconduct, then we consider it on a case-by-case basis and we consider it with reference to the core duties and the rules in the handbook. But where something has gone wrong, we would look at whether a barrister had taken steps to mitigate the situation."

This is where chambers governance becomes critical. Orlagh outlined what needs to be in place: "There's quite a broad church of things that you can do to help make sure that you're competent and using AI tools appropriately, and then also help demonstrate in the event something goes wrong or that you're questioned, that you have evidence of that. And some of these things, in my opinion, such as it is, are things that have not been asked of barristers individually before, such as drafting a business case."

Ewen added context: "The business case might be something that people don't think of initially. But I think it's a useful way of capturing what some of the risks might be and also thinking about how you'll use it, how it meets the problem that you're trying to identify."

Chambers responsibility is collective

The guidance expects chambers to think about AI strategy at a governance level. Ewen explained the collective responsibility: "There will be benefits in many cases of thinking about these things at a chambers wide level. I think from a chambers's perspective, it would be useful to have an understanding of how chambers's members are adopting new technology and what kind of risks that might generate for the chambers collectively."

The question for chambers directors and management committees is practical: do you know how your members are using AI, and have you thought through the risks? The guidance suggests specific steps.

The practical burden

Orlagh noted the operational reality for barristers and chambers: "There's quite a significant step change in the work that they would need to do...every time they use AI or every case that they manage, they have steps to take."

The BSB indicates that barristers should proactively disclose to their clients if they use AI in a way that materially impacts the delivery of legal services. But what does 'materially' mean?

When asked whether using AI to draft a cross-examination required client notification, Ewen was direct: "I would think so."

The question for chambers is which uses of AI require disclosure to clients and which don't. The guidance provides a framework but doesn't prescribe every scenario. Chambers need to work that out.

The competitive dimension

Orlagh raised a forward-looking concern: "There is possibly the chance that people who do decide to actively not use it over time will not be able to represent their clients to the best of their natural abilities because other people are using tools that are supporting them better."

Current state: 20% of reports involve AI

Ewen explained: "We and the other regulators are getting lots of reports about barristers which have clearly been drafted with the assistance of AI. I think we estimate probably about 20% of the reports that we get are produced with the assistance of AI."