MIT Technology Review reports that AI may not replace lawyers soon

MIT Technology Review reports that AI may not replace lawyers soon

MIT Technology Review’s article, published December 15, 2025, argues that AI is changing some legal work but may not replace lawyers anytime soon. It contrasts expectations raised by legal test results and automation forecasts with lawyers’ accounts of tools that struggle with case citations, ambiguous law and novel questions. The article argues that passing a standardized test is not the same as practicing law, which can require strategic judgment and arguments in unfamiliar situations. Professional benchmarks also find limitations, though the article notes that they may not capture the hardest work lawyers do. The case is not only about model capability: liability, regulation, clients wanting a human responsible, and firms’ reliance on billable hours may also affect deployment. The article reports that labor statistics had yet to show lawyers being displaced, while noting that AI may affect entry-level work and that future labor-market effects are uncertain. It also raises a training concern: Ethan Mollick says firms may need more formal teaching if junior work dries up. Junior associate Zach Couger values AI help with contracts but worries he is missing the repetitive practice senior attorneys received. He wonders whether he could be among the last class of associates at big law firms: “I may be the last plane out.”

Publication: 4f2feb08-b9ff-161d-f805-7a6538f72292

attribution — pass: The reading attributes the argument to MIT Technology Review, the publisher named in the evidence. No individual author appears in the text, and the reading names none. Quotations and views are correctly assigned to Ethan Mollick, Zach Couger and the other lawyers and researchers interviewed. The text is a single reported article, not a listing.

“I may be the last plane out.”

summary — pass: The title, summary and question faithfully convey the article's thesis: AI is reshaping routine legal work but is unlikely to replace lawyers soon. They also cover the gap between the bar exam and practice, benchmark limits, institutional barriers, labor data and the training concern. Couger's voice and the 'last plane out' line are preserved. The central term benchmark is defined on the page. Other terms, such as associate and billable hours, are understandable from context. The question follows directly from the Mollick and Couger material.

While AI is reshaping the grunt work of the profession, the end of lawyers may not be arriving anytime soon.

date — pass: The stated date, December 15, 2025, matches the UTC day of evidence.originalPublished and the URL path. The page was modified on 2026-01-05. However, every passage the reading relies on is plainly in the article body. The Deep Dive sidebar contains later-dated items, such as the AlphaGo ten-year piece, but the reading does not use or rely on them. Forecasts are reported as uncertain rather than as outcomes. This is a publisher page, so no archive disclosure is needed.

Capabilities may evolve over time, but that’s a big unknown.

support — pass: Each claim's excerpt is verbatim and supports the claim: routine use alongside complex-question failures, the bar exam versus practice, the 37% benchmark result, labor statistics, Mollick on formal teaching, and Couger on reps. The cited passages are plainly part of the original article body, so the later edit does not cast doubt on them. The summary's points on entry-level effects and institutional barriers are source-stated. The limitations correctly preserve the uncertainty.

Despite early signs that AI is beginning to affect entry-level workers, labor statistics have yet to show that lawyers are being displaced.

media — pass: No media was selected, and no verified media candidates were supplied. The null mediaId is appropriate.