Legal Geek 2024: answering the burning questions on Legal AI

At this year’s Legal Geek, we turned up the heat — literally.

Not only did our CEO Michael Grupp deliver his hottest takes on legal AI from the main stage, but we also spiced things during our Hot Ones-inspired workshop. If you didn’t get a seat early, you didn’t get in.

During our session, Playtime is Over. The Hot Truth About Legal AI, we asked the Legal Geek attendees to bring their toughest, most burning questions about the future of legal tech. No fluff, no dodging tough topics — just raw, honest discussion.

Our panel — featuring our CEO Michael Grupp, Nicola Shaver, CEO and Co-Founder of Legaltech Hub, and Richard Tromans, Founder of Artificial Lawyer — took on the kind of questions that usually go unasked at legal conferences.

Is legal AI just a marketing stunt? Are vendors overselling rebranded ChatGPT tools? What buzzwords are we all tired of hearing?

With so many hot questions, we couldn’t cover them all in one session, so we followed up with our panel to answer those we missed.

In this recap, we’re sharing both the highlights from the session and the insights we gathered post-event. Spoiler: It only gets hotter from here.

There’s no denying that the rise of AI in the legal sector has generated a lot of hype — but is it more than just marketing bravado?

Nicola Shaver weighs in on this, addressing both the real potential of AI and the challenge of sifting through marketing noise:

“No, legal AI has real efficacy and usefulness in the context of legal business and legal work, but unfortunately any period of hype gives rise to some marketing stunts and a frenzy of press releases that can mask the real work that is being done.”

Richard Tromans doesn’t mince words when it comes to the question of substance versus spectacle:

“No, there is some hype, but this is for real. It’s the most substantively powerful technology to engage with the legal world in decades.”

Meanwhile, Michael Grupp acknowledges that while real gains are still materializing, there’s no question about the technology’s genuine promise:

“No, the power is real. But it is true that there is more noise than productivity gains so far… Adoption rates may be lagging, but they are up.”

There’s a pervasive perception that every AI tool in legal tech is just a rebranded version of ChatGPT with a new label.

Answering the question directly, Nicola Shaver clarifies that while tools may use similar foundational models, the difference lies in how they’re built out for specific applications:

“No, but it can be hard to tell how much work has gone into developing a product built on GPT or another large language model to make it suitable for legal work. There is a misconception in the industry that many products out there are ‘mere wrappers’ around a foundation model. All wrappers are not the same, however.

There is a difference between thin and thick wrappers. Thin wrappers are effectively the deployment of a foundation model through API with minimal additional functionality added, whereas thick layers have a sophisticated application layer added to ensure the product is suitable for legal work. Buyers of legal tech need to be sophisticated enough to understand when a product is a thin wrapper and when it is more than that, and to evaluate which is needed for their particular use cases.”

Echoing this, Richard Tromans explains that the real value lies in the engineering and customization beyond the base model:

“All genAI tools will be leveraging an LLM of some type, but on top of that model or models is a lot of refinement, system prompting, workflows, and…engineering…that means these tools are a lot more than a rebrand.”

Michael Grupp sees the technology as a foundational layer to build upon, not an end in itself:

“We are building on the shoulders of giants. But there is much more to the new solutions than just ‘the giant’…work is needed to use Language Models for specific use cases.”

Every industry has its overplayed themes, and legal tech is no exception.

Richard Tromans cuts to the chase on one of his top irritants:

“That just using genAI in a limited way is a game changer just on its own — it’s only a game changer if you CHANGE THE WHOLE GAME.”

For Richard, true transformation means challenging entrenched systems, like the billable hour, not just layering AI on top of them.

Michael Grupp, on the other hand, sees enduring industry staples like the billable hour sticking around:

“The death of the billable hour. I am convinced there will always be a need for specialized, individual consulting… the core law firm business model will not change.”

While technology might be a game-changer, Grupp believes that some things — like the tailored expertise that clients value — are here to stay.

This debate heated up during the session, but while there was some disagreement over the future of the billable hour, there was broad agreement on the impact of AI.

The AI revolution isn’t just ushering in new tools; it’s also challenging the viability of established tech providers.

Nicola Shaver offers a broader perspective, noting that adaptability to technological shifts will be key for survival:

“It’s hard to tell whether there is a vertical that will not survive in its entirety, partly because AI allows companies to refresh and transform the way they solve a problem. No-code might have been an area in 2023 that people could have said would not survive, since coding using generative AI became feasible instead.

But if you look at the companies in that space, they have mostly thrived, through pulling in generative AI features and allowing for AI-enabled workflows. BRYTER now considers itself an AI company and sells agents. So instead of nominating a vertical I would say that any vendor that does not move fast enough to reimagine their technology or product in light of technological advances will not survive.”

Michael Grupp foresees a shakeout in heavily commoditized verticals:

“Tech that is already heavily commoditized may not survive. For example, document automation, document management or matter management… there will be a new generation of many categories. I think even very small current wrappers won’t stay. If it only takes a couple of months to build a solution, there is not enough defensible technology. But the new tech gives the opportunity to re-think a lot of the stuff and there will be a new generation of many categories where the incumbents could not adapt.”

In Michael’s view, technology that lacks defensible, unique value may struggle to survive, but with change comes the chance to reimagine and reinvent key tools in legal workflows.

What is the next technology tool after AI?

When it comes to future technologies, Nicola Shaver reframes the question, emphasizing AI’s continued evolution rather than speculating on a replacement:

“This is not the right question to be asking. AI is not just one thing; it’s a vast toolkit of methods and technologies that continue to evolve at pace. Instead, ask what the next hot area of legal AI is likely to be. My answer to that would be: agents and agentic AI. If 2023 was the year of large language models and chat bots, 2024 is the year of agents and workflows, and that will continue into 2025.”

For those wondering what could follow current AI technology, Michael offers a clear prediction: we’re only just beginning to see what AI can do for the legal industry.

“There will be iterations of AI for the next years. This is not a short wave. We have not even seen real adoption yet. We are only beginning to see the first vertical use cases emerging, so my prediction is that there will be several iterations of leveraging the power of (Gen)AI in the next years.”

Michael suggests that future applications will refine and deepen AI’s role in legal practice, where many workflows are only just beginning to see the transformative potential of generative AI.

What are some examples/use cases of AI transforming an organization?

Michael Grupp sees one of the most exciting use cases in what he calls “reading at scale.”

In his view, AI can democratize processes like due diligence, opening up access to tools previously reserved for large firms with the resources to get an ROI on heavy-duty software.

“Reading at scale is a very good use case of AI — and making diligence available for everyone. Until now, diligence software was only available for large organizations as it required sufficient data, training, and support through a team to really get to an ROI. That meant it was only useful for big projects. Now, the technology is available almost entirely out of the box. Solutions like BRYTER AnalyzeAI are more versatile, flexible, and affordable.”

AI is expanding possibilities for firms of all sizes, making advanced tools available to a broader range of organizations — and leveling the playing field in the process.

The heat doesn’t stop here

Legal Geek 2024 made one thing clear — the conversation around Legal AI is only getting hotter.

From the burning questions we tackled live to the ones we couldn’t fit into the session, Legal AI is evolving fast, and there’s no room for bluster. Whether it’s navigating the AI revolution or figuring out what comes next, the legal tech world is facing some big shifts, and we’re ready for it.

If you want to see the hottest Legal AI in action, book a demo here to see how you can use AI contract review software across a wider range of use cases. Let’s explore how AI can transform your legal workflows and keep you ahead of the curve.

Book a personalized demo