Future of law: firms seek to keep up with AI and the metaverse

Polly BotsfordMonday 8 January 2024

Mark Zuckerberg, Founder and CEO of Meta, once called virtual reality and the metaverse ‘the successor of the mobile internet’. The metaverse refers to an ‘interconnected world of extended reality’, navigated by, for example, virtual and augmented reality devices. By early 2022, land values in Decentraland – a popular metaverse – had reached a floor price of 5.24 Ethereum, the equivalent of a staggering $11,760 at current exchange rates.

But fast forward to late 2023 and the floor price of metaverses has fallen, while Reality Labs – the company developing Meta’s virtual reality, software, platforms and hardware – posted a loss of $3.7bn in Q3 of 2023. During the intervening time, of course, the generative artificial intelligence (AI) program ChatGPT was released and the conversation changed – although Zuckerberg said in October that the metaverse remained a ‘major long-term focus’ for his company, alongside AI. ‘ChatGPT and AI in general have eclipsed the metaverse as a priority for law firms,’ says Albert Agustinoy, formerly the Associations and Committees Liaison Officer of the IBA Technology Law Committee and a partner at Cuatrecasas in Barcelona.

Despite the arrival of generative AI, some firms have already made in-roads into the metaverse. ‘We have definitely got new clients through it, and some new recruits too’, says Eric Wagner, a partner at Gleiss Lutz in Stuttgart. ‘Legal advice is worth a lot if you have had practical experience of a problem, if you know the law but can apply it in a proper manner. Although the metaverse is very young, we felt that we needed to go through the whole process, what it feels like, what are the problem areas, so that we understand our clients’ struggles.’

But when OpenAI’s ChatGPT landed in late 2022, many firms’ focus shifted and they now take a ‘more cautious’ approach to the metaverse, says Agustinoy, not least because of the inevitable glitches and concerns around quality that come with new technology. ‘As a firm […] we didn’t find the platform […] to be mature enough to help us provide a satisfying experience. And you need the hardware to take that leap forward’, he says.

[We gave generative AI] a hypothetical scenario by asking for an analysis of the allegations of a counter party […] The outcome was far better than one might have imagined

Albert Agustinoy
Former Associations and Committees Liaison Officer, IBA Technology Law Committee

Generative AI programs such as ChatGPT can, in a law firm context, produce rather than simply review documents, for example. As such, they’re being harnessed by firms in various ways. The likes of Cuatrecasas and other large firms, such as Allen & Overy and Macfarlanes, for instance, have entered into relationships with Harvey, a large language model focusing on legal work. Addleshaw Goddard announced in November that it’s taking on a range of tools including CoCounsel, Spellbook and Microsoft Copilot, which it will engage with alongside its own internal generative AI platform. Hogan Lovells have incorporated all their legal tech into a separate brand, Eltemate, which includes machine learning.

Nicky Cranfield is Head of Employment at Ignition Law, a UK-based boutique firm for start-ups and scale-ups. She road tested ChatGPT, deciding to try an unwieldy Particulars of Claim (anonymised) and asking for a summary, which the program managed quite well, providing an outline timeline of the allegations. It also listed the main categories of claims that were being brought.

Cranfield says however that ChatGPT ‘missed some of the nuances of the claims.’ For example, ‘it captured most of the key events but some were not in the correct chronological order and a few described quite strangely.’ All in all, ‘it was absolutely astounding how quickly it threw out that information. ChatGPT provides a great starting point, and it can pull information together much more quickly than a human,’ she says. Cranfield adds however that the results need to be reviewed, tweaked and in some cases corrected in order to capture everything accurately.

Agustinoy is astonished at what generative AI can achieve. In litigation, for instance, his firm recently tested it in ‘a hypothetical scenario by asking for an analysis of the allegations of a counter party and what questions to ask witnesses in a trial. The outcome was far better than one might have imagined.’ But, as many such programs are designed to always provide an answer, ‘there are pitfalls,’ he says, such as where the AI invents pieces of case law rather than leaving the user without a response.

Despite ChatGPT’s prominence, it’s not necessarily a choice between generative AI and the metaverse: they are different elements in the tech landscape and law firms may wish to work on both. Aaron Grinhaus is Founder of Grinhaus Law in Canada and Director of the newly established Metaverse Bar Association (MetBA), a global network of lawyers providing development programmes and in-person and virtual networking. Less than a year old, the organisation already has 235 members. ‘The metaverse is the platform on which we all communicate and transact [while] generative AI is for content creation and automation,’ he says, highlighting that both require different legal attention. ‘The metaverse is not going away and it gives rise to novel legal issues that are prevalent in cross border, international commercial transactions and social interactions’, he explains.

Moritz Holm-Hadulla, a partner at Gleiss Lutz and Head of its Digital Economy Group, says that for firms hoping to engage with any technology, the advice is to ‘get involved at an early stage’. And someone, or a group, at the firm has to take the lead, as ‘it’s about individual effort because it is a lot of work.’ But, he adds, new technology needs to be considered across departments, sectors and so on, cutting across the silos that inevitably build up in any practice. ‘It’s also about having groups that are monitoring these trends and can come together to spot developments and then do something with it.’

Firms may need to buy in expertise, though Wagner says that Gleiss Lutz deliberately decided not to use external agencies deliberately ‘to have our own experience of building the programme’. They may also wish to implement a trial period before committing to roll out a technology to the wider firm. When Addleshaw Goddard was considering its generative AI strategy, it trialled 50 technologies before shortlisting the best five, which were then piloted with 150 staff across the firm.

Image credit: Oleksandr/AdobeStock.com