Lady Justice stands weighing and armed, even as the tools are quietly changing hands.
Image: Gaétan Marceau Caron on Unsplash.
Once upon a time, I was a huge fan of American legal dramas. The Practice, Boston Legal, and of course Suits, which I watched till the second or third season, around the point the storylines started to lose me and my love for TV began to ebb.
So it caught me slightly off guard to see Gabriel Macht, who played Suits’ hotshot closer Harvey Specter, turn up in my feed fronting a legal AI platform.
The platform is called Harvey, named after the character. Harvey signed Macht as its first brand partner earlier this year, and the clip in my feed was an instalment of its “Leading Questions” series: Macht in conversation with Harvey's co-founder Winston Weinberg about Suits, the practice of law, and the future of the profession.
A few thoughts, in no particular order.
First, I’m digging the meta-ness of this content-marketing play. Macht not bringing the fictional persona of Harvey Specter to bear is really powerful. Partly because it plays into the post-modern tensions of performed ‘authenticity’ on social media, and partly because I suspect there’s a major IP licensing premium attached to going the full-character route.
Second, there’s something deliciously meta in the premise itself. Harvey exists to augment legal expertise in ways that don’t require expensive, inhibitive credentialisation. So it’s rather fitting that Gabriel the skilled actor, rather than Harvey the brilliant lawyer, is the one who shows up. Nothing’s stopping them wheeling the character out later, once they’ve fully let us in on the behind-the-scenes conceit.
The campaign leads with, “The human side of lawyering will only become more important.” It’s a curious promise from a company selling a tool that automates lawyering.
Which is where my old teacher comes in.
Natasha Blycha taught me English at a boarding school outside Gweru, Zimbabwe, around the year 2000. She reignited a love for reading I’d lost. She’s now an emerging-technology lawyer, and when she joined me on the African Tech Roundup Podcast some months ago she argued the law as designed can’t cope with AI agents operating unsupervised.
Previously, on Karl Kruszelnicki’s Shirtloads of Science podcast, Blycha flagged that some courts had already found technology-assisted review more accurate than teams of junior lawyers sifting documents, and deemed it in the public interest to favour the machine.
In China, the Supreme People’s Court has written AI into the courtroom by policy.
Its 2022 opinion builds AI-assisted evidence review, similar-case retrieval and draft-judgment generation across the whole process of a case, and targets whole-process, high-level support across the court system by 2030. The same document reserves the final decision to the judge and treats the machine’s output as reference only. The court in one of China’s Tier-1 cities, Shenzhen, says its judges cleared 50% more cases last year with AI in the loop, and it’s exporting the system to dozens more cities.
The West appears to be heading in a similar direction by way of the client and the pipeline. Corporate legal departments now expect their outside firms to keep pace, and the gap is stark, with roughly 70% expecting productivity gains and only 6% believing most of their firms deliver them. At the same time, the entry rung is thinning.
Big Law’s model depends on junior associates learning on the job, and that is the work the tools swallow first. Paul, Weiss, Rifkind, Wharton & Garrison (Paul Weiss) embedded Harvey across its lawyers from 2023, and Clifford Chance cut jobs citing AI.
For context, these aren’t lookalike firms on either side of the pond. Clifford Chance is London Magic Circle (one of the UK's top five law firms), sporting thousands of lawyers across a global office network, with topline revenue in the billions.
Paul Weiss is a leaner Wall Street beast, a fraction of the headcount, but pulling far richer margins from high-stakes litigation and private equity.
These are firms running decidedly different models, and at least one of them is leaning rather bullishly into AI augmentation.I figure that once the bench runs on it, the clients demand it and rivals bill against it, adoption stops being a choice.
Unfortunately, none of this (necessarily) makes the machine trustworthy. Confident fabrication and fake case citations are wont to turn up in real court filings, to say nothing about the nagging institutionalisation of societal bias that’s always compromised legal structures the world over.
Macht says he apologises to the young fans who tell him they went to law school because of Suits. Partly, as I read it, for the false impression Harvey Specter gave them of how easily all the answers come… until tools like Harvey materialised it into something like reality.
That leaves me with two questions. Is the legal profession an early case of how long-held asymmetries of knowledge, skill and access are being systematically undermined?
And if the credential is no longer the moat it once was, what might it still be a ticket to?
Andile Masuku is co-founder and executive producer at African Tech Roundup. He serves as executive editor of Future in the Humanities (FITH), powered by the SA–UK Chair in the Digital Humanities at Wits University. Connect and engage with Andile on X (@MasukuAndile) and via LinkedIn.
Andile Masuku is co-founder and executive producer at African Tech Roundup. He serves as executive editor of Future in the Humanities (FITH), powered by the SA–UK Chair in the Digital Humanities at Wits University. Connect and engage with Andile on X (@MasukuAndile) and via LinkedIn.
Image: Supplied.
Follow Business Report on Facebook, X and on LinkedIn for the latest Business and tech news.