Guest post: AI did not win a legal case in England, but it is changing law as we know it

By Alex Cook, senior partner at Helix Law

It is time for a reality check. AI has not just won a court case in England.

The media, legal press and beyond, has been fizzing this week with sensational language: a “historic victory”; a “legal first”; even a case of “code beating counsel”, according to one magazine. And while we shouldn’t ignore that this is a significant moment for AI in our profession, it’s essential that we don’t get swept away by potentially dangerous hyperbole.

So let’s be clear again. No chatbots have won court cases in England this week.

I’m not anti-AI. In fact, I’m immensely excited by it. At our firm, Helix Law, we are incredibly proud of our achievements when it comes to developing and deploying technology to aid clients in our specialist litigation work. We were one of the first British firms to sign a contract with Harvey AI in 2024; AI tools have saved us over 6,000 man hours in the last two years; our technology does everything from AML checks and file opening, to transcribing phone calls and managing correspondence in disputes before the High Court. But ultimately, it’s humans – talented solicitors with years of experience – running the show.

In the recent Garfield AI case, software will have assisted in the recovery of the debt, helped prepare documents and streamline parts of the process. But much of the work at that level is process-based, rather than anything which requires a computer to ‘think’. Put another way: old-school template documents and mail merges can achieve very similar outcomes. Behind Garfield is a lone qualified solicitor, Philip Young, who is licensed by the Solicitors Regulation Authority; and in court, Garfield was represented by a barrister, Dominic Li.

The issue is less about Garfield marketing itself as AI-powered. Philip Young has every right to do so and plenty of other firms do too, including ours. But public perception matters, and when it’s widely reported that AI itself is fighting and winning court cases, the genie has escaped the bottle.

In May the SRA heralded Garfield AI as the “first purely-AI based firm” to be authorised in England and Wales. It is certainly not the only firm using AI and, as confirmed in the authorisation, Mr Young remains responsible for all work at the firm. The same SRA has also had a busy month rightly investigating solicitors and firms – in particular Pinsent Masons – over irresponsible or illegitimate use of AI in court proceedings. When the professional regulator puts individual firms on a pedestal for using AI in one breath, but warns and rebukes others using it with the next, it risks undermining confidence in firms and practitioners in the middle: those who know AI can have a transformative impact and are applying it properly, quietly and without fanfare.

There is a critical client trust question at stake here: with AI’s potential negative consequences coming to the fore, is now the right time to tell clients that human input can be removed from their legal services altogether? Do clients fully understand the limitations of AI-assisted responses? Would a client be happy knowing their case was being heard by an AI judge?

In the background, firms like ours have been investing in technology for years to improve efficiency and quality while reducing time spent on administrative work. AI makes it much faster to comb through disclosure, saves time wasted on internal emails (which we banned from our offices altogether last year) and frees us up to do what matters: apply human judgement to cases in order to win them.

The other big impact is on cost. As Philip Young rightly pointed out, integrating AI into case management can slash costs — we estimate by around 80% compared to City firms — making justice affordable and accessible. While Garfield AI’s case win was a positive, it was a simple debt recovery matter worth around £7,000. AI has helped us win cases issued in the tens of millions, for a fraction of the cost of a top London firm. Caseloads that once demanded significantly larger teams can now be managed far more efficiently, allowing firms like ours to compete with organisations many times our size. This is, we believe, the real historic change taking place.

Simply put, AI is not replacing lawyers, but it is making good lawyers more effective. It gives them more time to challenge evidence, refine legal arguments and build stronger cases. That is where clients see the benefit — and it is why we have continued investing just as heavily in our people, growing our headcount as we have in AI; something that cannot be said for larger firms keen to make savings on headcount without planting seeds for the next generation.

The Garfield case has shown that AI has a place within the legal profession. But I cannot see a future where it legitimately can be said to be conducting legal proceedings on its own. The firms that will lead the profession over the next decade will still be defined by the quality of their lawyers. The difference is that those lawyers will be equipped with the best tools and will know how to use them.