Do We Really Need ‘AI Regulation’?

In the US, House Republicans have added a proposed 10-year ban on regulating AI to a major piece of tax legislation. Meanwhile, the EU already wants to ‘re-do’ the AI Act, potentially watering it down. And the UK is going nowhere fast on making new laws for AI. But…what if that’s OK?

In fact, what if we don’t need to ‘regulate AI’? What if regulating something as amorphous as ‘AI’ is like trying to regulate the alphabet?

That is to say, we should be focused on the outputs first, then consider if there are laws that already cover those phenomena.

Here’s a couple of examples:

Copyright – now, it did really annoy me for a while when I found out that OpenAI had trained its LLMs on Artificial Lawyer without asking – along with most of the world’s websites…!

But, two things have mollified me somewhat since then: 1) OpenAI and other AI sites such as Perplexity are sending traffic to my website, and their inclusion of references to AL can be seen in a positive light, and 2) if you ask ChatGPT about an article in a newspaper or blog, it may summarise part of it, but it will say it cannot give you it all, and it will then point you to the source website (or at least that’s what I have seen).

In which case, is this an issue? If ABC Corp. LLM started to copy and then reproduce on demand entire articles verbatim from a media company, then that would be copyright infringement – pure and simple. You don’t need an ‘AI law’ for that. The same goes for music and books.

Now, again, does it pain me to see how someone has created a ‘phoney’ book that effectively copies the style and content of a famous author by using AI? Yes. But, again, if Mr ABC copies Harry Potter so closely that it looks like infringement, then you can bet traditional avenues will be open.

So…..does AI perhaps make people ‘lazy’ in terms of thinking and creativity? Yes, very much so, if one lets oneself go down that road. But, it’s hard to legislate against humanity’s taste for short-cuts.

Security and Facial Recognition – now, this was the one that got me supporting the EU AI Act when it first came out. I’ve never been a fan of the dystopias described in 1984 and Brave New World – (although clearly some see those books as ‘How To Guides’ for how to run a country, rather than warnings on what to avoid.) The ban on using AI for a whole range of things that could impose on our lives thus seemed very sensible.

But……..AI is just a tool, a method, to get the result, e.g. scanning everyone’s faces as they get on a bus even if they don’t want you to do that. Is that the AI at fault, or the government, city, bus company etc that allowed this?

Is the ‘AI’ the cause of this, or is it the bureaucrat who rather fancies the idea of being able to watch the population as if we were all just a bunch of zoo animals? I’d say the latter. The AI facial recognition is just the tool used to invade our lives, but it’s an organisation that decides to do such things. I.e. a law against organisations collecting facial data in general would suffice, and then there’s no need to pick on AI.

And we can see the same with many other areas. E.g. should we allow autonomous weapons? Well, they don’t need to be AI. Isn’t a landmine an autonomous weapon – you step on it, it kills you, no questions asked, no rules followed? Isn’t a squaddie with an itchy trigger finger while out and about without any rules of engagement also quite similar – and also a danger to all? Again, is it the tool, or is it someone making a decision that doing X in a certain way is OK that’s at fault?

I don’t claim to have the legal insights of a High Court Judge, but this is just what seems apparent from a common sense perspective. (And my thinking may well evolve again.) What do you think?

Any road, the debate about AI regulation is bound to run and run….until something even bigger than LLMs comes along….But, it does seem like we are perhaps approaching this without thinking through the practical realities of ‘regulating AI’.

—

P.S. re. the US amendment. The new text added to the tax Bill – which of course may not pass and hence will not become law – includes the below wording, which came from the Committee of Energy and Commerce:

‘Subsection (c) states that no state or political subdivision may enforce any law or regulation regulating artificial intelligence models, artificial intelligence systems, or automated decision systems during the 10-year period beginning on the date of the enactment of this Act.

Subsection (d) provides definitions for key terms used in the Act, including “artificial intelligence”, “artificial intelligence model”, “artificial intelligence system”, and “automated decision system”.’

And it’s worth saying that this definition of ‘AI’ is so broad it would include Google search and also my iPhone, which automatically makes many decisions for me, including dimming the screen after 10PM.

P.P.S. and when it comes to lawyers using AI…..again, isn’t there enough regulation there already? E.g. when a lawyer hands a client a legal document that they have written, isn’t there already a duty of care that they have their basic facts right? Where does it say in any Bar Rules that it’s ok to make things up? QED, the old rules still work even when we add AI to the mix. What we really need is more education on how to use these tools.

—

Richard Tromans, Founder, Artificial Lawyer, May 2025

—

Legal Innovators California Conference, San Francisco, June 11 + 12

If you’re interested in the cutting edge of legal AI and innovation – and where we are all heading – then come along to Legal Innovators California, in San Francisco, June 11 and 12, where speakers from the leading law firms, inhouse teams, and tech companies will be sharing their insights and experiences as to what is really happening.

We already have an incredible roster of companies to hear from. This includes: &AI, Legora, Harvey, StructureFlow, Ivo, Flatiron Law Group, PointOne, Centari, LexisNexis, eBrevia, Legatics, Knowable, Draftwise, newcode.AI, Riskaway, Aracor, SimpleClosure and more.

Cooley, Wilson Sonsini, Baker McKenzie, Gunderson, Ropes & Grey, A&O Shearman and many other leading law firms will also be taking part. 

See you all there!

More information and tickets here.