Artificial intelligence is rapidly becoming part of everyday business life.
It is helping organisations analyse information, improve efficiency and reduce administrative burden. Law firms are no different.
However, recent court decisions have highlighted a different side of the AI conversation. Not what happens when AI works well, but what happens when it isn’t properly checked.
When AI Goes Wrong
Two recent High Court decisions have attracted significant attention.
In Tobosaru v Court of Law Craiova, Romania, the court was required to address submissions containing legal authorities that did not exist. The Crown Prosecution Service subsequently accepted that the incorrect citations were likely to have originated through the use of generative AI. The court observed that increasing use of AI in legal work may be both necessary and beneficial, but highlighted the risks of relying on AI-generated material without appropriate oversight.
Similarly, in Anthony Malcolm Cork & Anor v Mark Smith, the court considered correspondence which referred to statutory wording that simply did not exist. Following investigation, it emerged that AI had been used during the preparation of the material and that the output had not been properly verified before being sent to the court. The judgment contains a detailed examination of how AI-generated errors can find their way into legal work if safeguards are not in place.
These cases are unusual, but the message from the courts is clear.
AI can be a useful tool; however, it is not a substitute for professional judgement.
What Does This Mean for Businesses Deploying AI?
AI is already being adopted by businesses across a wide range of sectors. Organisations are using it to increase efficiency, analyse large volumes of information, improve customer service and support decision-making.
Disputes often involve large volumes of information. Contracts, emails, financial records, project documents and internal communications can quickly accumulate into thousands of pages.
Technology can help legal teams review that material more efficiently and identify key issues earlier.
Equally, businesses may use AI before any dispute arises, whether to review contracts, support compliance processes, analyse operational data or assist with internal investigations.
Businesses considering how to adopt or expand their use of AI may also wish to consider the wider legal issues around artificial intelligence, including intellectual property protection, regulatory compliance, contract negotiation and workplace use.
However, the recent court decisions highlight a wider point that applies well beyond the legal profession.
Whether AI is being used to analyse documents, assist with internal investigations or support decision-making, the output should always be treated as exactly that: an output, not a conclusion.
It still needs to be assessed, challenged and verified by somebody with the experience to understand whether it is correct.
There is also a separate risk that businesses or individuals may seek to use AI as a substitute for legal advice when dealing with disputes. Whilst AI may assist with tasks such as summarising documents or identifying issues for further consideration, it cannot replace an understanding of the litigation process, procedural obligations, evidential requirements or the strategic decisions that often influence the outcome of a case.
Even where AI-generated output appears plausible, that does not mean it is legally accurate, complete or appropriate to the circumstances. Organisations and individuals should therefore view AI as a support tool rather than a substitute for professional legal advice, litigation strategy or informed professional judgement.
For businesses, that means AI adoption should not be treated purely as a technology project. It is also a governance, legal and risk management issue. Boards, senior leadership teams and governing bodies need to understand where AI is being used, what data is being processed, what decisions it may influence and what controls are in place to manage legal, regulatory, contractual and reputational risk.
For professional service firms in particular, proactive AI and data governance can help demonstrate accountability, manage compliance risk and support responsible innovation before issues arise.
This includes putting in place clear policies, approval processes, human oversight, audit trails, data protection checks, intellectual property safeguards and staff training. Without that framework, organisations may find that AI is being used informally across the business in ways that create unintended exposure.
A clear AI policy can help set the parameters for how staff use AI tools, what information can be shared, when human review is required and how legal, data protection, confidentiality and intellectual property risks should be managed.
The Opportunity as Well as the Risk
Much of the commentary surrounding AI focuses on instances where things have gone wrong.
When deployed appropriately, AI can help organisations process information more quickly, identify trends that might otherwise be missed and free up time for higher-value work. In the context of disputes, it can assist with document review, chronology building and identifying potentially relevant issues within large datasets. However, those benefits should not be overstated. If AI is used without appropriate governance, it can also create legal, regulatory, data protection, confidentiality, intellectual property and reputational risks. The opportunity therefore sits alongside the need for clear policies, human oversight, approval processes and accountability.
The lesson from the recent court decisions is not that AI should be avoided. Rather, it is that technology delivers the greatest value when combined with proper human oversight and expertise.
Our Approach
At Stephens Scown, we recognise the benefits that technology can bring.
Used appropriately, AI can improve efficiency, assist with document analysis and allow lawyers to spend more time focusing on strategy, advice and achieving the right commercial outcome for clients.
What it cannot do is replace legal judgement.
The recent court decisions demonstrate why that distinction matters. The courts did not criticise the use of AI itself. The issue was that the results were relied upon without being verified.
For us, technology is a tool, not a decision-maker – any use of AI must sit alongside professional expertise, critical thinking and supervision.
We also advise businesses more generally on how to deploy AI safely and responsibly. That includes helping organisations understand the legal risks, develop practical governance frameworks and put in place policies and controls that reflect how AI is actually being used within the business.
Our work in this area can include AI policies, staff guidance, data protection and privacy assessments, contractual protections, intellectual property risk management, procurement support and board or leadership training. The aim is not to slow innovation down, but to help organisations adopt AI in a way that is lawful, controlled and commercially useful.
This reflects the wider guidance we have published on artificial intelligence, proactive governance in the AI and data era, and the importance of putting an AI policy in place. Those themes are particularly important for businesses whose governing bodies need to show that AI use is being considered at an appropriate level, with clear accountability and proportionate controls.
Businesses that want to explore this further may find our related guidance helpful, including our materials on artificial intelligence, proactive governance in the AI and data era, and why every business should put an AI policy in place.
Looking Ahead
AI is not going away, nor should it.
The legal sector, like every other industry, will continue to explore ways in which technology can improve efficiency and client service.
The recent cases are not a warning against innovation. Instead, they reinforce the importance of responsibility, accountability and professional judgement when using powerful new technologies.
For businesses involved in disputes, that should provide reassurance. The most effective outcomes are likely to come from combining the efficiencies offered by technology with the judgement, experience and accountability that people provide. AI can help improve the process, but the responsibility for the decision-making remains firmly with those using it.
This article was written by Harry Cooke, Trainee Solicitor, and Jeremy Crook, Partner in the Commercial Dispute Resolution team, with contributions from Amy Ralston, Partner in our Intellectual Property, Data Protection and Technology team.