AI in Legal Services: Risks, Rewards and the New AI Sensibilities
This blog series by Peter Impey, Consultant General Counsel at InTouch, looks at what AI is, how it’s evolving, and what it means for legal services. It outlines key developments and use cases, touches on risks, regulation and the SRA’s position, and explains InTouch’s practical, low-risk approach to bringing AI into everyday work in a law firm. You can find the full series and related articles on the InTouch blog.
In 2025 the rise of AI was hard to ignore. On 11 December 2025 a press release from the UK Government announced Google DeepMind as a partner. On offer, according to the release, are potentially incredible developments including “clean energy for the future”, “smarter public services”, “AI in education” and “responsible AI development”.
These rewards are massive, if the promises and potential can deliver actual outcomes. If you are at all still sceptical about the scale of impending change, here is what some significant voices are saying about the AI revolution:
“AI in recent years has been underhyped.” — Geoffrey Hinton
“AI is so profound, its influence is hard to overstate.” — Bill Gates
“AI and robotics is a ‘supersonic tsunami’, so it is really going to be the most radical change that we’ve ever seen.” — Elon Musk
Jeff Bezos says AI is a “horizontal enabling layer”. In his view the biggest impact will be that “it will affect every company in the world”. Quality and productivity will go up. “It is real, but hard to understand.”
With tech titans saying such things we have to stop and think. Take their words seriously and believe that everything around us, including the fabric of society itself, is going to change.
What about legal services?
Given all this, it is absolutely right to consider the risks and rewards of AI in legal services. Not only are the skills we use in daily life changing, but entirely new companies and services are being created because of the rapid advances in AI.
However, early in the LLM hype cycle there were a number of well-publicised horror stories in legal. Lawyers, for example, were fooled by hallucination errors into making false case citations or filing AI-generated defences in US courts.
The bad publicity was widespread and put people off, but much has changed and there are now significant rewards on offer. The benefits to personal and professional productivity have accelerated, and LLMs have shifted from novelty experimentation to providing integrated assistance in day-to-day work.
Horror stories quickly appeared
Many of the difficulties came from a lack of understanding. A mistaken belief that an LLM is a fact-based knowledge tool or a supercharged search engine failed to recognise the risks of hallucination. This was often made worse by lawyers covering up their misunderstanding and lying to the court. Sadly the penalties were harsh: suspension from practice, fines, public humiliation and professional embarrassment, and so on.
While individuals grabbed the headlines, quieter horror stories of other significant failures involved data leakages, privacy breaches and billing errors. Once again, a lack of understanding of the technical characteristics of LLMs was a root cause.
Law firms and even practice management systems failed to realise that client and confidential data loaded to an LLM was exposed to external companies. Worse, the information was reused as part of the ongoing development of the models.
Confidential case summaries uploaded into AI-enabled practice management tools for review or manipulation were sucked into the model as training material. Worse still, if they loaded their policy documents and precedent materials, firms were donating the intellectual property that is the foundation of their practice.
Some horrors from this early period may not even be exposed until a contract is litigated in the future. Firms that used AI assistance may have baked in clause failures to AI-negotiated contracts. Or their fast, AI-enabled contract reviews may have missed negative covenants in M&A deals, for example.
Mistakes like these are ticking time bombs primed by inaccurate contractual work or sloppy due diligence on M&A deals done during the early months of the AI boom. These examples show the risks involved in deploying new technologies within legal practices.
Accordingly, law firms wishing to deploy AI should adopt a calculated risk-acceptance approach. Focusing on risk mitigation rather than avoidance means you can benefit rather than miss out. The upside potential is dramatically bigger than the downside of ignoring the paradigm shift taking place in legal services that I talked about in an earlier blog.
Using a “closed-loop” system like InTouch prevents issues with sending data externally, and security and privacy are uppermost in mind at InTouch.
Rewards in legal from using AI
The promise is that well-delivered and thoughtfully used AI will improve:
Turnaround times
Service levels
Insights and controls
Risk management
Cost management
Workflows
Knowledge management
There is commercial value in all of these aspects that is impossible to ignore. Accuracy and speed enhance the delivery of legal services. AI should be used to help with productivity, precision, quality of service and affordability.
Clever use of AI at the right point in a workflow makes the most of the technology to improve the outcome. On InTouch you can ask for assistance at any point in the matter lifecycle. Data remains on the system and a full audit trail is kept.
Efficiency and innovation for cost reduction
AI can reduce time for tasks like contract review and create innovative service solutions. It is especially powerful where workflows and tasks are repetitive from client to client.
New client applications such as automated client query resolution or first-line advisory can improve the client experience at low cost to a law firm. In the next blog I’ll explain more about the use cases already live on InTouch.
Enhanced operations and reporting capabilities
AI can assist with the creation, curation and review of materials and matters. Contextualised text and document generation capabilities can help support staff with invoice processing, or lawyers by generating matter summaries.
By answering “where are we?” questions with insights, lawyers can provide updates to clients quickly and team members can much more easily step into projects to help out as needed.
By using AI for analysing and synthesising data, managers can identify trends and make management decisions with greater confidence. Resource allocation and hiring decisions become easier, especially if the AI is itself considered part of the resource mix.
Focus on client-centric services
AI helps with scale, transparency and innovation. It enables owners to grow their practice and keep down the cost base while delivering superior legal services. It is even possible to use AI in the marketing, sales and publicity aspects of a business.
How to claim the rewards?
Moving beyond the fear of the horror story stage to understanding the benefits and rewards on offer is part of a maturity pathway. Legal professionals who take an interest and are prepared to innovate are building a set of AI sensibilities to enable a disciplined deployment of AI.
Initial scepticism and concern is being replaced by rigorous understanding and risk control. Decision-makers now have a much more sophisticated understanding of the underlying technology of LLMs and best practices for reducing failure points (like hallucination) and retaining control.
Lawyers must retain responsibility for data, especially confidential client information. Keeping the lawyer in the loop, as happens on InTouch, and establishing working practices and firm policies is normal.
Following the initial hype when horror stories emerged, the legal industry reacted strongly with bans and fear. But some innovators started to pull ahead and build dedicated teams and resources. Large firms and in-house teams started with back-office tasks and gained a head start in the rewards race.
With controlled integration of AI, a law firm can move past the horror and, by adopting new sensibilities, start to create a sustainable competitive advantage.
Next in the series, we look at InTouch’s practical, low-risk approach to AI, and how to bring it into everyday work in a law firm so lawyers and support teams can benefit safely and at scale.
This series of blogs about AI is written by Peter Impey, General Counsel for InTouch. If you’re exploring how to use AI in your firm, you can find the full series and related articles on the InTouch blog.
