How Are Lawyers And Law Firms Using AI? And What Are The Risks?

Traditionally, the legal sector has not been well known for its embrace of technology. In fact, for a long while it’s been considered fairly prehistoric in its approach to digitalisation. However, the legal industry has been uncharacteristically quick to adopt AI.

Whilst 76% of senior legal leaders reported no AI use in their organisation in 2024, now, in 2026, only 2% can make that claim. Deloitte’s report, The AI Imperative: Reshaping of the Legal Industry, also tells us that 61% of legal departments are now deploying AI, and collectively, amongst those surveyed, AI is expected to save or automate an average of 28% of legal work over the next two to three years. Being that AI is capable of automating workflows and assisting with or even managing day-to-day administrative tasks, it may be unsurprising that the legal industry is moving ahead with adopting AI to ease the workload. After all, such tasks have kept lawyers away from casework and contribute to the high-stress nature of the profession. However, the law sector also has a reputation for being meticulous, valuing accuracy and being hyper-aware of the risks of misinformation and errors. Just recently (August 2026), a lawyer in the UK became the first to be struck off for using AI-generated citations (that proved to be inaccurate) in a tribunal. In the US, a trial was also cancelled, and lawyers on both sides were sanctioned after it was discovered they had both cited AI-hallucinated cases in court. AI in law is famously known for hallucinating and creating fake case law that we could use as a precedent, with fake citations and it’s happened in the UK too. In fact, the High Court has urged UK lawyers to take action to prevent AI misuse, following several cases being impacted by the use of artificial intelligence that presented courts with inaccurate or hallucinated evidence. So, is the risk worth it? In this article, we explore how lawyers are using AI in their practices, how they can mitigate risk and how AI might reshape the legal profession in the next few years.

How is the Legal Sector using AI?

The 2026 Actionable Insights for Law Firm Leaders Report, by Thomson Reuters, found that the legal sector is not just using AI for back-office tasks. 77% of those surveyed were using it for document review, 74% for legal research, 74% to summarise documents, and 59% were using AI tools to draft briefs and/or memos. Interestingly, they also found that use of AI was higher in almost every area in the legal sector than in government. Other reported uses of AI in the legal sector included -

• Contract drafting

• Correspondence

• Knowledge management

• Question answering services

Whereas AI tools are being far less utilised for marketing, business development, human resources, risk assessment, finance and ESG reporting. So, in terms of AI application, it’s very much weighted toward assisting with actual legal work and less on supporting other business functions. These reports have been somewhat surprising since the legal industry has been understandably protective regarding the accuracy of any documents, research, or contracts that might be used in court or by a client. However, AI has come a long way in a short time, and AI tools have been created specifically for law firms and lawyers. Perhaps because AI creators recognise the problem the legal profession has with burnout. With 92% of lawyers having faced job-related stress or burnout, and over 25% experiencing this daily (YouGov survey), maybe it’s no wonder that a sector known for being slow on the uptake of technology historically has decided to take a different approach with AI.

So what are the benefits AI is providing to the legal sector?

How are Lawyers Benefiting from AI?

Lawyers are using AI to streamline workflows - for time tracking, automating billing, client communications and task management. However, AI is also being used for document review, research, contract drafting, summarising, and document processing. During the COVID-19 enquiry, Norton Rose Fulbright used AI-assisted e-discovery tools to process thousands of documents. In large cases like this, AI can significantly speed up processes, leading to faster outcomes and arguably more informed decision-making, since it can process more documentation in less time. Smaller law firms and independent lawyers are also finding AI tools useful. With fewer resources, time-consuming tasks impact how many clients they can manage. Therefore, lawyers are harnessing the powers of AI to draft contracts under their instruction, often doing so using their own templates and requesting AI to adjust, as necessary, to individual client needs. Generative AI can already do this, but agentic AI takes it to a new level, often asking further questions to better complete the task, or identifying risks or scenarios specific to the client that may have been overlooked. Furthermore, AI may be able to extend capabilities. Its global legal knowledge, for example, could enable small local firms to take on clients needing to comply with laws in other continents or countries. Or it may offer insights into a sector that a lawyer less experienced in that industry may not have known to consider. Hence, AI could help smaller firms grow faster and compete.

Not all lawyers or law firms are at this point of their journey with AI though. Some are less trusting - “If I’m doing legal research on a novel topic, I might use Google at the very start of the process, and there is an AI element in the response I get. After that, I will always, always read the actual original source (legislation or case law) before I would think of committing an argument to paper.” (reddit forum comment)

The Risks of Using AI in Your Law Firm

AI Inaccuracy And Bias

AI in legal research is arguably one of the most contentious uses. Time saved in using AI for research is considerable, but is it trustworthy? After all, AI isn’t trained only on screened and fact-checked data; it’s evolving all the time, mining the World Wide Web for information (and we know the internet is rife with misinformation and bias), and it can hallucinate. What does that mean? It means sometimes AI will totally make things up; like a child wanting to tell the most interesting story to please their audience, it has a tendency to falsify information.

What’s important to understand is that there aren’t just some AI tools capable of producing inaccurate or biased information - an AI tool can be completely reliable most of the time and still make an error randomly. This is why complete oversight and meticulous fact-checking are essential for any AI-generated work and when using research carried out by AI.

Data Protection

Feeding client data into an AI tool poses a confidentiality risk. Many public models use the information a user inputs to train themselves and, even if this setting is turned off, sharing confidential materials with AI (case files, contracts, or personal data) is effectively sharing it with a third party. This can result in the potential loss of the information’s privileged status. With no clear legal framework specifically governing AI in the UK, and many law firms lacking AI usage policies, legal professionals may be inputting sensitive data into AI tools without fully understanding how those tools handle it. One overlooked risk, for example, is that AI tools may process information outside of the UK and EU, which can breach UK GDPR. To mitigate the data protection risk of using AI -

• Opt for legal-specific AI tools (these tend to be more secure)

• Ensure UK hosting, non-sharing of data for training purposes and other data protection guarantees

• Practice data minimisation; only input what is necessary, when it’s necessary, and anonymise and redact where possible

Accountability

Accountability for AI-assisted work sits with the lawyer, not the tool. The SRA (Solicitors Regulation Authority) has been clear that the “AI got it wrong" is not a defence, and solicitors remain fully liable for errors that reach a client or the court. Professional indemnity insurance, though, and internal sign-off processes are also having to catch up with where the responsibility actually lies when an AI mistake slips through. With the buck stopping with the individual lawyer, mitigating risk becomes essential. That starts with understanding how a given AI tool actually works, so its weaknesses and potential failure points can be identified before they cause a problem. Firms, and even sole practitioners, should be drawing up AI policies that set out how AI will and won't be used, what can and can't be input, and how human review is built into any client-facing work.

Transparency is also part of being accountable. Many businesses using AI are now beginning to disclose this in their client contracts, and lawyers should be too. Be clear on what AI is being used for and whether personal data is being shared. Explain how the firm will protect information and ensure expert human oversight.

Should AI be used in the Legal Sector?

AI provides many benefits to the legal sector and can free up time for lawyers to focus their expertise where most required. It also has the potential to speed up processes, meaning improved efficiency for both firms and their clients. However, AI should not be a reason for lawyers to move to the passenger seat. Legal work will always require an expert at the wheel. Therefore, when integrating AI into the workplace, leaders must be sure to understand the risks, implement processes to mitigate them, and ensure that all AI outputs are not only given ‘the once-over’ but fully analysed for accuracy. That way, law firms can ensure they’re part of an AI success story, and not a cautionary tale.