What are the risks of using AI for law firms?

Whilst artificial intelligence (AI) presents huge opportunities for the legal sector, and is becoming ever more prevalent, professionals must remain mindful of their own regulatory responsibilities. We have seen regulatory action and judicial criticism of lawyers both within England & Wales and in other jurisdictions, following incidents such as hallucinated legislation and cases, fake case law before the High Court and the Employment Tribunal, and witness statements drafted with AI in the Family Court.

Pinsent Masons is the latest firm to be criticised for providing references which featured AI hallucinations, and then for using AI to produce an explanatory letter that was still wrong (and contained no apology to the Court for the errors in the first incorrect submission). The firm has since referred itself to the Solicitors Regulation Authority.

Legal professionals must ensure that they are up to date with recent guidance published by regulators. We now have 2026 guidance and compliance tips from the Law Society’s AI Strategy, SRA and BSB guidance as well as CILEx resources. As is so often the case with regulation, when something goes awry, professionals will need to be able to evidence a documented risk-based approach, appropriate supervision, and an awareness of guidelines.

Responsibility stays with the lawyer

The key point is that, regardless of technology, from a regulatory standpoint the lawyer will always remain the person making the decisions and with professional responsibility. A professional must be able to explain and justify their own work. It is crucial to ensure proper supervision of the work of junior colleagues as well as awareness of their use of AI.

Core ethical and conduct principles which may well be engaged by the use of AI are the requirements:

• to maintain client confidentiality,
• not to mislead the court (knowingly or recklessly),
• to act effectively with proper governance and risk management, and
• to act with honesty and integrity.

Practical compliance tips

Some tips to bear in mind:

• Put AI in your governance framework: keep the use of AI part of a firm’s governance controls and under COLP/Compliance review with suitable monitoring and risk-based controls in place and properly recorded within policies and addressed by training.
• Train staff and keep within your competency: ensure training on technology to maintain a sufficient level of competence, and awareness of how AI may be used by opponents and clients.
• Focus on data privacy and confidentiality: consider your GDPR and other data processing obligations.
• Enable choice and transparency: allow clients to make informed choices as to how their matter is handled and be open if asked by others about the use of AI.
• Advise caution: be very cautious in the use of AI for preparation of witness statements where concerns with its use may lead to the weight of that statement being limited, i.e. where a court is unclear as to whether this is in fact the independent recollection of the witness.
• Verify: check and check again with an eye for AI-generated errors, omissions, or hallucinations – it is safer to treat it as an aid. Carefully review AI output, verify it, check it for completeness and accuracy, and adapt it to your work rather than solely relying on it.

If you have questions or concerns about your firm’s use of AI or regulatory obligations, please contact Hannah Pilkington.