Skip to content
Worthclock

Blog / 6 min read

The AI Your Law Firm Never Chose: How to Check Built-In AI Features

By Shang Ahmed · 26 October 2026

Built-in AI in email, documents and meeting tools can expose client data. How law firms can check which AI features are switched on.

Quick Answer

Many law firms already use AI without ever choosing a tool. AI drafting, summarising and transcription features now come built into email clients, document editors, meeting software and browser extensions, and some are switched on by default or enabled by individual users. Because nobody installs anything, these features rarely go through any review. For a law firm, that matters: European and UK guidance says client information should only go into AI systems with proper safeguards in place. The fix is a short inventory of which AI features are active in the software you already pay for, followed by a decision on which ones may handle client data.

What Are Built-In AI Features in Law Firm Software?

When firms talk about AI risk, they usually picture a lawyer pasting a document into a public chatbot. That risk is real. But a growing share of AI use in firms looks different. It arrives inside tools the firm already uses:

  • Email, where AI can draft replies, summarise long threads or suggest responses
  • Document editors, where AI can rewrite, summarise or extract key points
  • Meeting software, where AI can record, transcribe and summarise client calls
  • Browser extensions, where AI assistants can read whatever page or document is open

None of these requires a purchasing decision. Someone clicks "try it", or the feature appears in an update. The firm never asks the questions it would ask about a new vendor, the same diligence questions before adopting any legal AI tool.

Can Built-In AI Breach Client Confidentiality?

The rules do not distinguish between a tool the firm bought on purpose and a feature that appeared in an update.

The CCBE, which represents bars and law societies across Europe, published a guide on generative AI in October 2025. It says lawyers should not enter client related data into generative AI tools unless safeguards are in place, such as contractual confidentiality terms, zero data retention or a data protection agreement, and that lawyers should understand how the provider uses what they enter.

In England and Wales, the SRA's warning notice of 17 August 2026 goes further. It says client information should only be entered into AI systems where appropriate contractual, technical and organisational safeguards are in place, that both free and paid AI systems may pose risks to confidentiality, and that legal professional privilege may be permanently waived.

The Upper Tribunal's decision in UK v Secretary of State for the Home Department [2026] UKUT 81 (IAC) shows how ordinary the trigger can be. It concerned client emails and Home Office decision letters put into ChatGPT to improve and summarise them. The Tribunal said that putting such material into an open AI tool places it in the public domain, breaching confidentiality and waiving privilege.

Email threads, call recordings and draft documents are exactly the kind of material built-in AI features process.

Why Do Law Firms Miss Built-In AI?

The 8am 2026 Legal Industry Report found that only 9% of legal professionals say their firm has a written AI policy that is actively enforced, and 43% say there is no policy and no plan for one. If most firms have not written rules for the AI tools they know about, very few have looked at the AI features they do not know about. (See also: Does Your Law Firm Have an AI Strategy? Most Don't.)

There are also practical reasons:

  • Features are enabled per user. One person turning on a meeting summariser can mean every client call they host is processed by it.
  • Updates change defaults. A tool that had no AI features last year may have several today.
  • Terms differ by plan. The same feature can come with different data terms depending on whether the firm has a consumer, business or enterprise plan.

How Can a Law Firm Check Which AI Features Are Switched On?

This does not need an IT project. A small firm can do it in an afternoon.

1. List the software you already use

Email, calendar, document storage, word processing, video calls, practice management and any browser extensions staff have installed.

2. Check each one for AI features

Look in the admin settings and recent release notes for anything described as AI, copilot, assistant, summary, transcription or smart reply. Note whether each feature is on for everyone, on for some users or off.

3. Ask the people who use them

Admin settings do not show everything. Ask staff directly which AI features they use, including browser extensions and meeting note tools.

4. Read the data terms for each active feature

For each one, find out whether your inputs are used to train models, how long data is kept, where it is processed and whether a data processing agreement covers it. If you cannot find clear answers, treat the feature as unsuitable for client information. Our guide on using AI with privileged client information covers what good data terms look like.

5. Decide and write it down

For each feature, decide whether it is allowed for client data, allowed only for non-client work, or switched off. Record the decision and the date, and tell staff.

6. Repeat after major updates

Put a short review in the calendar every quarter, or whenever a key tool announces new AI features.

FAQ

Is built-in AI safer than a standalone chatbot?

Not automatically. It depends on the terms of your plan and how the feature handles data. Some business plans offer strong data protections. Others do not. The SRA notes that paid tools may also pose risks.

Should we just switch all AI features off?

Usually not. Many of these features save real time, especially on email and meeting notes. The goal is to know which ones are running and make a deliberate decision about each, rather than banning everything.

Who should own this check?

Someone with authority over both IT and compliance. In a small firm that is often a managing partner. In England and Wales it fits naturally with the COLP's role.

Does this apply to firms outside the UK?

Yes. The SRA notice applies to England and Wales, but the CCBE guidance applies across Europe, and confidentiality and data protection duties apply to law firms everywhere in the EU.

Related reading

The Bottom Line

The AI most likely to put client information at risk is often not the tool a firm chose. It is the feature nobody noticed being switched on. A one afternoon inventory of the software you already use, followed by a written decision for each AI feature, closes most of that gap.

Not sure which AI features are already running in your firm, or which ones are worth keeping? Our AI Tools Assessment maps the tools and features you use and gives you a clear, safe plan for each.

Book Your Assessment

Sources