A consequence: AI demand for judgment costs firms the associate path

Latitude Legal named Alex Su its president, LawNext podcast reported on September 16, as the flexible legal staffing company described rising demand for lawyers with real judgment in an AI-driven workflow. In the same account, Latitude said it had seen a 52 percent jump in new law firm and corporate clients in the first half of the year. Su told host Bob Ambrogi that better AI would increase demand for lawyers who can direct it, check it and own the work.

That points toward the advantage that matters more than simply removing beginners from matters: firms need a way to produce capable lawyers faster. AI can reduce routine work, but the resulting model still depends on people who can exercise judgment over the finished work. The commercial signal in Latitude’s account is therefore compatible with a leaner legal workforce that values senior judgment. It does not, by itself, show that firms are developing that judgment faster.

The strongest objection is blunt. Jordan Furlong wrote that firms will not train associates after AI and will instead train partners. That is a coherent description of a profession abandoning the associate path rather than improving it. The answer cannot be that AI makes the old pipeline unnecessary. The long-term advantage belongs to firms that use AI to develop capable lawyers faster, not merely to firms that hire fewer beginners. A firm that removes junior exposure without replacing it has cut a cost, not built a capability.

The event proves only that Latitude was named a president and that Latitude reported a 52 percent increase in new clients during the first half of the year. It does not prove that flexible staffing produces better lawyers, that AI caused the increase, or that any firm has shortened the time required to develop legal judgment. A company’s account of its own client growth is not a performance result. The stronger reading would require matter-level proof that lawyers working with AI reach capable, independently checked output faster than lawyers in the old model. No such comparison appears here.

Other announcements show that the profession is responding to the training problem, but not that it has solved it. Non-Billable reported on September 29 that Legora was expanding its AI training programme to UK law schools and had signed its first university partners. Thomson Reuters Institute published a call to redesign new attorney onboarding for the AI era. Those are signs of institutional adjustment, not evidence of faster lawyer development. The next confirming fact would be a reported cohort outcome: the work trainees performed, the supervision they received, and later performance measured against a comparable path. Until firms publish that link, the associate path may be shrinking faster than a replacement for it is being built.