An Everlaw survey says in-house teams are already sending less work to outside counsel, Legora raised $550 million, Womble launched an AI lab for repeatable work, and the Law Society reminded lawyers that responsibility stays with them.
Everlaw and Wakefield Research surveyed 300 US in-house legal professionals between June 26 and July 30 and published the results on October 20. Of the departments that use outside counsel, 54% say AI has already reduced the volume of work they send, and 86% of all respondents use generative AI for legal work. Among heavy users the cut-back figure is 73%.
The most useful number is about price, not adoption. Asked what clients should primarily pay outside counsel for, respondents put business and commercial insight, complex matters and specialist expertise at 54–56%. Document review and routine legal work drew 28% and 27%.
Treat it as a vendor-sponsored survey of self-reported views. It stops at attitudes: it gives no spend figures, so it cannot say how much money has left outside counsel. Savings sit on the modest side too. Of respondents whose firms use generative AI, 92% report AI-related savings, but only 13% call them significant.
This is the problem statement in a respondent's own words. Few in-house teams think routine work is what outside counsel should be paid for, yet inexpensive work is still done by expensive resources because the work has to go somewhere. Using AI inside the legal team trims some of it. Outsourcing that work to supervised agents moves it off the lawyer's desk, and the firm's, entirely.
Legora announced a $550 million Series E on October 21, led by Atreides Management, at an $8.8 billion valuation. The company says annual recurring revenue now exceeds $200 million and that the money will go towards expanding across North America. TechCrunch notes that Harvey reportedly raised $550 million at a $15.5 billion valuation a month earlier, which keeps Legora second on valuation.
A funding round shows investors expect the market to grow. It does not show what a customer's legal team got out of it. Revenue growth measures software bought, and says nothing about how much work stopped reaching outside counsel or how often the output needed correcting.
When a vendor raises at this scale, who is the product built for: your lawyers as users, or the work as the thing being done? Ask for outcomes per workflow, such as turnaround, escalation rate and error rate on standard agreements, before you read the valuation as evidence.
The transatlantic firm, with 1,300 lawyers across 42 offices, announced Womble AI Labs on October 20. It is led by Chief AI and Technology Officer Chad Ergun and sits inside the firm's IT function. Artificial Lawyer reports that it targets legal and administrative tasks that are "structured, repeatable, and well-defined."
A firm describing a category of work as structured and repeatable is describing, in effect, the work its clients have been questioning the price of. The announcement does not say whether tools built by the lab will change what clients are billed for that work, or whether the savings stay with the firm.
Firms are building AI for the same repeatable work, which is useful confirmation that it is commodity work. But a lab inside the firm leaves the routing where it was: the work still arrives at the firm, on the firm's terms. Outsourcing legal work to supervised agents puts the sorting on the client's side, with the client's templates and escalation rules, and sends the firm only what needs a firm.
The Law Society of England and Wales published a practice note on October 21 as part of voluntary guidelines for legal AI. Its message is that "AI is not a legal entity," so responsibility for the work stays with the solicitor. It asks firms to choose tools carefully, check that work meets professional standards and apply human oversight to AI output.
The note also says AI is no substitute for junior lawyers, while accepting that some junior tasks can be replaced if suitable checks come before work reaches clients. It is aimed at law firms, but the principle applies to in-house teams too. A preliminary Law.com benchmarking read on October 16 found 43% of respondents' legal departments have no formal position on outside counsel using AI on their matters.
If responsibility cannot move to the tool, it has to sit with a named reviewer somewhere. Whether the work is done by your team, a firm or an agent, who is that person, and what must they see before the work leaves?
A survey says in-house teams are sending less to outside counsel, a vendor is valued at $8.8 billion, a firm is building its own lab, and the Law Society says responsibility stays human. The work is moving. The accountability has not moved with it.
Each story this week shows the same pattern from a different angle: the work is repeatable, the people doing it are expensive, and the tooling is sold to the lawyer who still has to supervise it. Inexpensive work is done by expensive resources until someone owns the routing.
Outsource legal work to supervised agents. Flank's agents know your templates, terms and escalation rules, take the NDAs, standard redlines and procurement agreements, and hand finished work to a named lawyer for review before anything leaves.