Structuring Counsel’s Instructions with Help from AI

Good instructions to counsel are still one of the most powerful documents in litigation.

Done well, they:

• organise messy facts into a coherent story;
• crystallise the issues the court actually has to decide; and
• ask focused, answerable questions.

Done badly, they leave counsel spending billable hours figuring out what is going on — or, worse, missing something important.

AI will not write counsel’s advice for you. It can help you structure the brief, surface gaps and assemble supporting material so that your instructions are clearer, faster to prepare and easier to update.

This article looks at how UK firms can use AI to structure counsel’s instructions while keeping solicitors firmly in control of content and tone.

1. Decide what your standard structure should be

Most sets of instructions benefit from a common spine, for example:

• Background and parties;
• Procedural history;
• Factual narrative;
• Issues on liability and quantum (where relevant);
• Evidence available or to be obtained;
• Instructions and specific questions;
• Timetable and practical constraints;
• Enclosures and key documents.

AI works best when it knows the structure you want. A good starting point is:

• agree a house template for counsel’s instructions in your practice area;
• capture it as a precedent or template inside your case management system; and
• let AI help fill the sections, rather than inventing new shapes each time.

2. Using AI to tame the facts

The hardest part of many briefs is turning:

• sprawling email threads;
• overlapping witness accounts; and
• transactional documents

into a clean, fair factual narrative.

AI can assist by:

• summarising long email chains against specific questions (“What has the defendant said about delay?”);
• extracting events from documents and notes into a chronology you can then refine;
• grouping facts by theme (background, negotiations, alleged breaches, post‑breach conduct).

A typical pattern is:

1. Link key documents and emails to the matter in your system.
2. Ask AI to create a draft factual summary or chronology table from those materials.
3. Edit that draft heavily, adding nuance, caveats and references.

The rule is simple: if you would not sign it yourself, do not send it to counsel. AI is a starting point, not a substitute for reading the file.

3. Clarifying the issues and the questions

Many instructions blur together:

• what has happened;
• what the client wants; and
• what you are actually asking counsel to do.

AI can help you separate these strands, for example by:

• taking a rough “brain‑dump” note from the fee‑earner and splitting it into:

  • facts;
  • possible causes of action / defences;
  • procedural options;
  • specific questions;

• turning a list of questions into numbered, tightly‑phrased issues (“On the assumption that X, is there a realistic prospect of…”).

This makes the final instructions easier to read and reduces the risk of counsel answering the wrong question.

4. Managing enclosures and document lists

Another perennial problem is enclosures:

• duplicated documents;
• missing key items;
• no clear mapping between the narrative and the bundle.

AI can support you by:

• generating a draft enclosure list from the matter file (titles, dates, short descriptions);
• suggesting groupings (pleadings, correspondence, contracts, expert material);
• pointing out where the narrative refers to a document that does not appear in the list.

Inside a system like OrdoLux, you can go further:

• link each enclosure to an ID in your DMS or bundle;
• have AI check that each document referenced in the instructions appears somewhere in the list;
• avoid the “mystery exhibit” that only exists in someone’s inbox.

Final responsibility for what is enclosed still lies with the solicitor — but AI can make omissions easier to spot before the brief goes out.

5. Tone, privilege and what AI should not touch

Instructions to counsel are often privileged and may contain:

• candid assessments of witnesses and strategy;
• internal disagreements within the team or client; and
• predictions about settlement or publicity risks.

When using AI around counsel’s instructions:

• keep processing inside approved, governed tools (for example, your case management system rather than public chatbots);
• minimise how much of your own evaluative commentary you feed into the model — AI does not need to see every internal aside;
• never ask AI to decide what advice you should seek or what strategy to adopt.

A good rule of thumb:

• AI can help organise facts, documents and questions;
• only lawyers should express judgment, impressions and recommendations.

6. Updating instructions as the case evolves

In many matters, counsel’s instructions are not a one‑off document. They evolve as:

• pleadings change;
• new evidence arrives; and
• interim hearings reshape the issues.

AI can be useful in:

• generating a marked‑up “delta summary” of what has changed since the last conference;
• updating chronologies and enclosure lists as new material comes in;
• helping you prepare short supplemental instructions rather than redrafting from scratch.

This is especially valuable in long‑running disputes where everyone is at risk of losing the thread.

OrdoLux is legal case management software built for UK solicitors, with AI tools integrated directly into the matter workspace.

Where OrdoLux fits

OrdoLux is a legal case management platform for UK solicitors. It includes a built-in AI legal research tool for case law and legislation research, with citations for human verification.

The platform handles matter management, time recording (via keyboard, automatic Outlook email capture, and WhatsApp), document storage with SharePoint, billing, KYC via Checkboard, Stripe payments, and electronic signatures — all in one place.

See all features or book a demo.