AI for in-house legal teams

The requests come from everywhere. Sales needs a customer's paper turned round before quarter end. Procurement has a supplier who wants the DPA back by Friday. HR needs an offer letter changed, and someone in marketing has signed up to a tool nobody has read the terms of. None of it arrives as a tidy matter. It arrives in Slack, in email and in the corridor.

It is mostly the same contracts. NDAs, supplier agreements, customer terms, DPAs, renewals. The positions barely change from one to the next. The value is in applying them every time, quickly, without missing the clause that is different.

The team is small. Often one GC, sometimes a handful of lawyers, occasionally nobody with legal in their title at all, just a CFO or COO who looks after the contracts. Holidays and sick days don't pause the requests.

Someone signs the budget. Outside counsel is expensive, headcount is hard to win, and every tool has to show it gives hours back. The legal team is judged on speed to the business and on the risks it catches, not on hours billed.

Good AI for in-house counsel starts from that shape. It has to take requests where the business already asks, apply one set of positions the same way every time, keep working when people are away, and keep track of what was signed.

• Intake from Slack. Link the channel the business already uses for legal requests to your legal team. Its envoys pick up what lands there, start a routine if one is set, and post back with your sign-off. People outside legal never need a new tool or a form. More
• Contract review against one playbook. Your positions, fallbacks and escalation points go into the envoys' memory once. Every contract comes back as tracked changes in Word, a comment on every change, your fallbacks applied and a summary of what's off-playbook. The GC's playbook is applied by every envoy, the same way. More
• Obligations and renewals. Envoys pull renewal and notice dates, payment and reporting duties, termination windows and who owns each one into a tracker. A routine checks it on schedule and tells you what's coming, before the notice period closes. More
• Due diligence. Load the data room, and envoys read every document, index it, pull out the key terms and flag issues with the page they came from. Several envoys can take parts of the room at once. More
• Records. Every contract an envoy touches is filed with its terms in the team's datastore, a structured store the envoys design and keep current. People read it; envoys keep it correct. More
• Reporting and your own CLM. Ask for a contract tracker, a renewals dashboard or a review screen, and an envoy builds it on your datastore. It is as current as your records, and you change it by asking. More
• Cover when someone is away. The envoys know the playbook, the people and the history. When a lawyer is on holiday, the work keeps moving and nothing waits for their return.

Contract review. In-house teams hand Claren their NDAs, supplier contracts, customer paper and distribution agreements, and get tracked-change redlines back against their playbook. The same positions apply to every contract, with a comment on every change.

Due diligence. Through an acquisition, a whole data room is read, indexed and reviewed. Every file is accounted for, and the issues come back in a form the deal team can use.

Renewals and obligations. Every signed contract and its terms go into the datastore. Renewals are flagged before they roll, and obligations are tracked without anyone keeping a spreadsheet.

Intake and cover. Requests from the business are picked up from Slack, from financial documents and HR policies to the long tail of small questions. The work carries on through absences and holidays, and the team gets hours back every week.

Most of the legal AI on the market was designed for law firms first. A firm's work is organised by matter and by client, billed by the hour, and done by lawyers who have the time to drive a research and drafting toolkit. The tools follow that shape.

Harvey describes itself as built for the world's top law firms and in-house legal teams. Its core is Vault for storing and bulk-analysing documents, Knowledge for legal, regulatory and tax research, and Spaces for working with co-counsel behind ethical walls. Its intake and work orchestration modules for in-house teams are marked new. It suits law firms that want a large research and drafting toolkit and have the lawyers free to drive it.

Legora sells to firms and in-house teams alike. Its Tabular Review pulls the same answers from many documents into one grid, and its Portal is for collaborating with clients. It is suited to bulk review across a matter. For in-house work that arrives one request at a time, it is the wrong shape.

Spellbook works mainly inside Microsoft Word, on contracts only, with playbooks, and an intake and renewal-monitoring module. It suits individual lawyers who want drafting help inside Word.

Each of these makes a lawyer faster at a task. An in-house team's problem is different: there are more tasks than lawyers. What it needs is the task taken, done and handed back, by something that already knows the team's positions and keeps the team's records. That is why Claren is built as AI employees, not as a toolkit. See how Claren compares with Harvey and Legora.

1. Does it do the work, or help you do it? A draft you still have to check, finish, file and send saves minutes. A contract handed back redlined, filed and tracked saves the afternoon.
2. Where do requests reach it? If the business has to learn a new portal, it won't. The work should be picked up where people already ask, which for most companies is Slack.
3. Is there one playbook, or one per lawyer? A team's positions are the team's. They should go in once and be applied the same way, whoever is on holiday.
4. Does it work when nobody asks? Renewals, notice periods and new contracts in a channel don't wait for a prompt. Look for work that starts on a schedule or a message.
5. Who keeps the records? Every contract reviewed should end up somewhere you can find it, with its terms and dates, without anyone doing the filing.

1. Choose the first stream. NDAs, supplier contracts or renewals: whatever fills the week and follows the same positions every time.
2. Give the envoys your playbook. Your positions and fallbacks go into memory once. No written playbook? Teach them as you go. See contract playbooks.
3. Link your requests channel. Connect the Slack channel the business already uses, so requests reach the legal team's envoys where they're already asked.
4. Set the routines. Review every contract that lands in the channel. Check the renewals tracker every morning. Each run is kept, so you can see what was done.
5. Keep the sign-off. Envoys ask before anything leaves the team. Read the first weeks closely, correct what needs correcting, and loosen the sign-off where you trust the result.

Claren holds a SOC 2 Type II report and ISO/IEC 27001:2022 certification. Data is encrypted at rest with AES-256 and in transit with TLS 1.2 or higher, and stored in US, UK or EU data centres. Data Processing Agreements are available on request.

All data is treated with privilege in mind. Each organisation's data is kept apart, nothing crosses between matters, and every user action is kept in an audit trail. Inside the company, teams draw the lines: legal's envoys, files and records are legal's, and a private team's work is visible to its members only.

Neither Claren nor its model providers train AI models on your data, a commitment formalised through zero-data-retention agreements with OpenAI and Anthropic. The Enterprise plan adds zero data retention and enterprise SSO for the whole organisation.

And the lawyer stays accountable. An envoy works on its own, but before it posts in Slack, uses a connected tool, messages another envoy or sends anything outside the team, it asks. You decide where it needs your sign-off, and where it needn't.

A solo GC can start on the Free plan. Pro is $70 per seat per month for one to three people, with $50 of usage per seat included. Team is $250 per seat per month for three or more, with $200 of usage per seat pooled across the team and shared team envoys. Enterprise is priced with you. See pricing.