How Can Law Firms Manage AI Risks in Legal Services?

Law firms need stronger governance frameworks to manage the growing use of artificial intelligence in legal services, according to a Law Society feature published on 18 November 2026. The article brings together the perspective of Peter Lee, partner and global head of AI governance at Simmons & Simmons, on how practices can benefit from AI while managing professional, operational and regulatory risks.

Why does AI governance matter for law firms?

The increasing capabilities of AI systems are creating new opportunities for legal practices, including potential productivity improvements and the ability to automate aspects of work. However, those opportunities must be considered alongside the duties that continue to apply to solicitors.

The Law Society feature emphasises that regulators in England and Wales, together with judicial guidance, have made clear that solicitors remain responsible for their work even when AI is involved. Professional responsibility cannot be transferred to an AI system.

Firms therefore need to consider more than whether a particular tool can perform a task. Their approach should also address the reliability of AI-generated outputs, the protection of legal privilege, client confidentiality, data protection and cyber security. The use of increasingly capable or agentic AI systems adds further questions about the extent to which technology can take actions on a user’s behalf.

AI adoption needs to be supported by clear governance, informed decision-making and appropriate human oversight.

Summary of the governance themes discussed in The Law Society feature

The Law Society’s AI practice note is identified as a useful starting point. However, each firm must translate general guidance into policies and controls that reflect its own services, systems, people and risk profile.

How can firms build an effective AI culture?

AI knowledge and confidence can vary considerably across a legal practice. Some employees may already be familiar with the technology, while others may be uncertain about how it works or when it is appropriate to use it.

The feature recommends regular and adaptable training rather than relying on occasional one-off sessions. As AI tools develop quickly, training and internal guidance should also be reviewed frequently so that staff have access to current information.

Firms should make it possible for employees at different levels to raise concerns, share experiences and suggest improvements. Listening to feedback from junior colleagues as well as senior staff can help practices identify practical issues that might otherwise be missed.

Training is also relevant to professional competence. The article notes that firms within the scope of the EU AI Act may also need to consider the Act’s AI literacy obligation.

What is shadow AI and why is it a concern?

Shadow AI refers to the use of AI tools by employees without the approval or oversight of their organisation. It may happen when a firm has not provided suitable tools, when staff do not understand the risks, or when employees bring familiar consumer applications into their working routines.

The concern is particularly serious when staff enter client-confidential information or personal data into public or consumer AI services. Depending on the provider’s terms and practices, information submitted to a tool may be stored, reviewed by provider personnel or used to train future models. Once information has left a firm’s controlled systems, retrieving it may not be practically possible.

Such use can create risks for confidentiality, legal professional privilege and data protection compliance. The Law Society feature suggests that firms can reduce the incentive for unsanctioned use by making approved tools accessible and straightforward to use, supported by a proportionate approval process.

It may not be appropriate or affordable for every practice to invest in a specialist legal AI platform. Nevertheless, firms should understand which tools their staff are already using and establish a clear route for assessing and approving technology.

What should practice leaders do?

Firm leaders are encouraged to create an AI roadmap that sets out how the organisation will assess tools, implement policies, govern use and train employees. A documented approach can help reduce avoidable risks to a firm’s reputation and professional standing while giving staff greater clarity about what is permitted.

A practical first step is to prepare an inventory or register of the AI systems and features already being used across the practice. The level of scrutiny can then be matched to the risks associated with each use case, rather than applying the same controls to every tool or activity.

  • Map current use: identify AI tools and AI-enabled features already used by teams and individuals.
  • Assess risk: consider the nature of the task, information involved, potential impact and level of human oversight required.
  • Set policy: explain which tools are approved, what information may be entered and which uses require additional review.
  • Train staff: provide regular guidance on responsible use, limitations and professional obligations.
  • Review regularly: update controls as technology, firm needs and relevant regulatory expectations develop.

The feature also links good governance with client engagement and business development. Where staff understand the firm’s approach, AI may be used more confidently as a support for professional judgement and the quality of legal work.

How can standards and collaboration help?

International standards and risk frameworks can provide useful reference points for organisations developing AI governance arrangements. The feature highlights ISO/IEC 42001, the International Organization for Standardization’s AI management system standard, and the National Institute of Standards and Technology’s AI Risk Management Framework.

These frameworks can help firms consider how AI systems are introduced, managed and monitored. The article encourages legal professionals to engage with developments in technical standards and best practice as well as changes to formal law and regulation.

It also identifies an opportunity for legal practices, regulators, law societies and universities to work together. Academic research can bring together expertise from computer science, law, philosophy, humanities and other disciplines, helping the profession examine complex questions that firms may not have sufficient time or resources to investigate independently.

The Law Society says it engages with government, regulators and other stakeholders on issues including AI regulation, data protection, automated decision-making, copyright and the use of AI in courts. It also encourages members to share their practical experiences so that its policy work and guidance reflect the needs of the profession.

Subscribe to our newsletter

Frequently asked questions

Are solicitors responsible for work produced using AI?

Yes. The Law Society feature states that solicitors remain responsible for their work whether or not AI is used. Responsibility cannot be passed to an AI tool.

What is shadow AI in a law firm?

Shadow AI is the use of AI tools or AI-enabled features without the firm’s approval or oversight. It can include employees using consumer AI services for work tasks without considering the confidentiality and data protection implications.

What should a law firm include in an AI policy?

The feature points to clear governance, approved tools, staff training and policies for assessing AI systems. Firms should tailor their controls to their own circumstances and the risks of each use case.

Why is entering client information into public AI tools risky?

Depending on the provider’s terms, submitted information may be stored, reviewed or used to train future models. This can place client confidentiality, legal privilege and data protection obligations at risk.

Which AI governance frameworks are mentioned?

The feature refers to ISO/IEC 42001 and the NIST AI Risk Management Framework as useful sources for understanding approaches to AI governance.

Does every law firm need to buy a specialist legal AI platform?

The feature does not suggest that one platform is suitable for every firm. It highlights the value of providing appropriate, approved tools and having proportionate processes for assessing technology.

Don’t Miss Key Legal Updates

Get SRA rule changes, SDT decisions, and legal industry news straight to your inbox.
Latest news
Related news