Open navigation
Search

Using AI in Legal Practice: Takeaways for In-House Counsel

27 Oct 2026 United Kingdom 2 min read

Key contacts

As AI becomes part of everyday legal practice, understanding the professional and regulatory risks is becoming just as important as understanding the technology itself.

Last week The Law Society of England & Wales published a practice note on the use of artificial intelligence in legal practice. The note brings together much of the existing guidance and case law on AI into a single resource and is applicable across the profession, including in-house legal teams. With an estimated reading time of over 1 hour and a multitude of links to further resources, this is a reference document to bookmark.

While necessarily high-level in approach given the breadth of the legal profession, it provides an important update to the framework for considering the legal, regulatory and professional issues arising as AI becomes increasingly embedded in legal practice and the wider organisations that in-house teams support. Coming soon after SRA | Misuse of AI - Warning notice | Solicitors Regulation Authority, it is essential reading across the legal profession.

Here are six takeaways for in-house counsel from the practice note:

  1. AI governance is a leadership issue: Accountability sits with leadership. Communicate acceptable use and ensure adequate training, even if AI isn't formally adopted.
  2. Know what AI you are actually using: Watch for shadow AI, modifications or updates to AI tools and AI features quietly added to existing software.
  3. “Enterprise-level” is not a guarantee of protection: Check contracts, privacy terms and settings in terms of data handling on each system.
  4. Protect confidentiality and privilege: Understand whether your AI includes adequate protections for confidential data. Privilege over prompts and AI logs remains unsettled. Be cautious.
  5. Agentic AI needs tighter controls: Limit permissions, set human approval points, name responsible individuals for verifying outputs.
  6. You remain responsible, not the AI: Verify against primary sources. Using AI to check AI won't meet your duties. Be aware of possible hallucinations, and discriminatory or biased outputs.

Read the full practice note here: Using artificial intelligence in legal practice | The Law Society

Back to top Back to top
Opens in new window