Artificial intelligence (AI) is starting to play a more visible role in legal practice. As these tools become more capable, the key issue is not whether they will be used, but how. In particular, where do we draw the line between the use of AI as a tool and the provision of legal advice for which a solicitor remains responsible?
In practice, the distinction is straightforward. AI can be used to support internal tasks such as research, summarising material and preparing draft documents without difficulty. The position changes, however, once its output is used to shape advice or is shared externally with clients or third parties. At that point, it falls within the solicitor’s professional responsibility and must be treated accordingly. The line is therefore not defined by the technology itself, but by whether the output has been subject to proper scrutiny and independent legal judgment before it is relied upon.
Case law demonstrates that the risks associated with this shift are no longer theoretical. In R (Ayinde) v London Borough of Haringey [2025] EWHC 1040 (Admin), the High Court was confronted with a situation in which legal submissions included multiple non-existent authorities. The court found that several cited cases were entirely fabricated and described their inclusion as “wholly improper” and professionally unacceptable. Notably, the judge considered that such conduct could amount to professional misconduct and made a wasted costs order against both counsel and solicitors.
The approach in Ayinde is reinforced by a more recent tribunal authority in R (Munir) v Secretary of State for the Home Department [2026] UKUT 81. The Upper Tribunal considered submissions which relied on authorities that were, in substance, the product of AI “hallucination”. The Tribunal treated this not as an isolated or technical error, but as a failure going to the heart of professional competence and the duty owed to the court.
Courts have shown little tolerance for reliance on AI outputs that have not been independently verified. The risk is not limited to obvious inaccuracies. AI systems are capable of producing material that appears legally coherent but is substantively wrong, whether through fabricated authorities, misapplied principles, or outdated law. For solicitors, this creates a heightened duty of scrutiny. The use of AI does not simply introduce efficiency; it introduces a new category of risk that must be actively managed.
There is a temptation to treat AI purely as a risk to be controlled, but that would be too narrow a view. The same technology that has led to the issues identified in Ayinde and Munir also has the potential to enhance the quality, accessibility and efficiency of legal services in ways that were not previously possible. The challenge for the profession is not whether to adopt AI, but how to do so in a way that strengthens rather than undermines professional standards.
Firms are integrating advanced platforms such as Harvey AI, Ivo and LegalFly into their workflows. These tools are designed specifically for legal use, enabling lawyers to process large volumes of material, identify relevant authorities, and generate structured drafts in a fraction of the time traditionally required. When deployed within a framework of proper supervision and verification, such systems can enhance both productivity and the overall quality of output.
AI is already making legal information more accessible to individuals and businesses who might otherwise struggle to obtain timely advice. Properly calibrated, AI tools can act as a gateway, supporting clients in understanding their position before engaging solicitors for tailored advice. While these tools may assist in providing a general overview, they cannot replace the careful analysis, verification and professional judgment required to advise on a client’s position. The risks associated with so-called “AI hallucinations” make it essential that legal advice is obtained from qualified and regulated professionals, rather than inferred from automated systems.
As technology continues to evolve, firms that adopt a structured approach are likely to be best placed to benefit. Those that view AI solely through the lens of risk may find themselves at a competitive disadvantage. AI will form part of the future of legal practice. The task for solicitors is to ensure that its use enhances, rather than erodes, the trust placed in the profession.
Our firm is well placed to provide guidance to ensure compliance in these areas. Please do not hesitate to contact us if we can be of assistance.
