AI led complaints in the disciplinary context: access or anathema to justice?

In his keynote address to Oxford Civil Justice Systems in the 21st Century Conference in May 2026 titled “AI and Civil Justice: Preparing for the Tsunami”, Lord Briggs warned that the courts would see a “tsunami” of small to medium value civil claims from litigants in person caused by the use of generative AI.

While the speech was about claims, at least in that context the rules on costs (certainly in theory) stand as a (potential) guard against abuse of the system. But what about in the disciplinary arena where = when considering complaints to regulators – in most contexts the costs guardian has yet to be deployed?

In our experience, complaints to regulators have in recent years been increasingly used as a “stalking horse” to test whether there is a potential claim against a regulator. Does then the lack of costs sanctions provide a backdoor route to avoid the protections faced in the civil arena? And if so, is now the time for a change?

Are regulatory complaints increasing? 

In our experience, complaints to regulators have in recent years been increasingly used as a “stalking horse” to test whether there is a potential claim against a regulator. Does then the lack of costs sanctions provide a backdoor route to avoid the protections faced in the civil arena? And if so, is now the time for a change?

By some reporting at least, for the final quarter of 2025, the Solicitors Regulation Authority (SRA) reported that the number of complaints it received was 80% higher than the same period in 2024. Similarly, the Bar Standards Board also noted that complaints about its barristers had increased by 27%. Complaints to the Legal Ombudsman were also 37% higher in late 2025 as compared with the same period in 2024.

This mirrors our experience, not just in those particular contexts, but in disciplinary actions more widely. AI is, almost certainly in part, driving this increase. A complainant can now produce a complaint and supporting documents within a matter of minutes and these can, on their face, appear compelling. Not only that, but some of the complaints we are seeing are demonstrably lengthier than those we would tend to see not only increasing the demand on defending them as a matter of number, but also in terms of time needed for any one dispute (and with that, of course, costs).

How generative AI can assist regulatory complaints 

That technology can be used in this context should not of course be seen to be entirely a bad thing.

The disciplinary sphere exists – at least in part – to ensure that professionals who should be held to account, can be. That technology – and specifically generative AI – can be used to allow people to do this without incurring significant costs must be considered something which in theory at least is compatible with this regime.

Particularly given that every disciplinary context is unique – and the rules in some can be complex and at time far less developed than say in the civil context – means that in terms of preparation and process, generative AI offers potential obvious gains. If that then avoid the situation whereby a potential complainant is not then deterred from making what would be a valid complaint due to the amount of time or the technical skills required to prepare the relevant documents and in researching and following the correct processes, at a macro level that is positive.

The risk of AI-generated disciplinary complaints 

On the other hand, if it takes only a matter of minutes, there is a risk that complaints become very easy to make and that regulators become overwhelmed by AI-fuelled complaints, some of which would have no real prospect of succeeding and which are unnecessarily lengthy. There is also the risk that these complaints contain “hallucinations”, errors made by the AI in its drafting of the documents that mean that the complaints are not reliable.

Incidentally, those acting in defence of disciplinary proceeding should not necessarily see this as entirely negative to their role as – certainly in our experience – often the use of AI makes it easier to establish, challenge and dismiss the frivolous claims, and we have a strong track record in that regard. This is however not what the disciplinary process should be seeking to achieve.

Should costs sanctions apply in disciplinary proceedings?

This of course all comes at a cost for those defending disciplinary actions and indeed for regulators, many of whom have stretched capacity already. Whilst the ‘ship has probably sailed’ in terms of whether or not generative AI will be used, this – alongside a wider trend of more aggressive regulator action (reflecting wider political trends) does to us give good cause to ask again whether now is the moment to bring the costs guardian back into the disciplinary process? No doubt doing so could go a long way to avoid AI being used to abuse the process.

This would of course require many regulators to rethink their processes and procedures, but that is probably something that needs to be done in any event. Given the ever-developing costs regime in the civil context, this would not be breaking new ground, rather a redeployment of the already trained costs guardian to a new assignment.