Skip to content
BPP, home page

AI in legal practice

The tool can produce the work. The lawyer must be able to stand behind it. 

Artificial intelligence is already changing the way lawyers work, with legal AI tools supporting research, drafting, document review, due diligence and case analysis. But it can't take responsibility for legal advice. Understand how AI in law can be used responsibly, the risks involved and the professional judgement needed to use it effectively. 

Two people reviewing work on laptops at a shared table in a contemporary office, with large windows and workstations in the background.

Where AI is being used in legal work

AI is one of the most significant developments in legal technology and is already helping legal professionals work more efficiently across a wide range of tasks. 

Research and analysisLegal research, case analysis, chronology building and document summarisation.
Drafting and reviewFirst draft preparation, document review, due diligence and comparison against precedent.
Knowledge and workflowKnowledge management, matter triage and workflow automation within legal teams.

The SRA describes AI tools as having ‘become commonplace in the workplace and our everyday lives’ and notes that many solicitors and firms use them safely and responsibly. 

Being used doesn't mean being relied on without oversight. Understanding where AI helps is only part of the picture. Understanding where human judgement remains essential is just as important.

What can’t be transferred to the AI tool 

Four responsibilities remain yours, whatever produced the draft. 

1. The decision 

You must still decide if the output is accurate, appropriate and fit for purpose. An AI-generated response is a starting point. Deciding whether it’s right requires professional judgement. 

2. Your professional duties 

Your duties to the client, the court and the proper administration of justice remain unchanged. Regulatory obligations don’t disappear because an AI tool was involved. 

3. The safeguards 

Confidentiality, privilege and appropriate supervision remain essential, including when reviewing AI-assisted work produced by others. 

4. The explanation 

You must be able to explain why the final work is accurate, appropriate and safe to rely on. 

What the regulators say

Solicitors - SRA

Client interests, supervision, accuracy 

The Solicitors Regulation Authority (SRA) states that AI has no separate legal personality, and that solicitors using it remain accountable for their work and outputs, regardless of how that work was produced. Reliance on an AI-generated output is not a defence. Firms are expected to have appropriate governance, systems and controls in place to manage the risk, and those supervising others remain accountable for work carried out under their supervision. 

From the SRA warning notice, Misuse of AI, published 17 August 2026. 

‎

Barristers - BSB

Duties to the court, independence, verification 

The Bar Standards Board (BSB) applies its existing professional duties to AI use. It expects barristers to demonstrate a basic level of technology and AI competence, verify output before relying on them, and take a risk-based approach when adopting any tool. The BSB also warns against using free generative AI tools where terms permit providers to retain submitted information. Professional independence cannot be delegated. 

From the BSB guidance on artificial intelligence, published May 2026. 

Understanding AI risks in legal practice

These risks aren’t hypothetical. They’re the reasons regulators have issued warnings about the use of AI in legal work. 

A group meeting between staff and students across a large table

Fabricated authorities

Tools can generate fictitious cases and citations that read convincingly and have no basis in fact.

Confidentiality breaches

Entering client information into public tools can create risks around confidentiality, privilege and data protection.
Two students sit across a desk infront of large windows during the day

Plausible but wrong

Incorrect propositions, outdated law, missing context and fluent language can be mistaken for accuracy.

These risks have already impacted real legal proceedings.  

In R (Ayinde) v Haringey LBC [2025] EWHC 1383 (Admin), fabricated citations were placed before the court. The solicitor and barrister faced a wasted costs application and were referred to their regulators.  

In UK v Secretary of State for the Home Department [2026] UKUT 81 (IAC) the Upper Tribunal commented on the potential for the use of public AI tools to breach confidentiality and privilege. 

The risk isn't that AI can be used. The risk is relying on it without proper verification, supervision and professional judgement. 

Develop the expertise to perform when it matters

From law conversion courses and SQE preparation to Bar training and apprenticeships, explore the routes into law and build the knowledge, skills and judgement needed for legal practice. 

Speak to a law expert