Table of Contents

• EDITOR’S RAMBLE 🗣

• FEATURED REPORT 📰

• IN OTHER NEWS 🗞

• AROUND THE WEB 🌐

• STUFF THAT MIGHT HELP YOU 👌

EDITOR’S RAMBLE 🗣

I’m on holiday this week, so I’m writing to you from a sunny beach in Rhodes.

But as I always say, three things in life are certain:

• death,

• taxes, and

• LittleLaw in your inbox (a day late this time — blame it on the beach).

- Idin

FEATURED REPORT 📰

📚 This US ruling just reshaped AI copyright law

What’s going on here?

A US court has ruled that Anthropic, the AI company behind Claude (a ChatGPT competitor) can legally train its AI models on copyrighted books as long as they were purchased legitimately.

This marks the first major victory for using the “fair use” doctrine for AI training.

Why is this decision important?

Courts are trying to strike an important (and difficult) balance between innovation and protecting creators' rights.

💻️ For AI companies: This ruling is big — before this, companies didn't know if using copyrighted books to train their AI models would lead to lawsuits. Even books they had bought legally were risky to use.

The court decided that training AI on legally bought books counts as “fair use”. This sets a strong precedent in favour of AI-focused tech companies in the US.

📚️ For authors and publishers: They argue that AI companies “steal” their work without permission or payment. So, the decision is a blow to authors and publishers. But the decision doesn't give AI companies total freedom — they are not allowed to use pirated books. So, they’ll still need to legitimately buy any books they want to use to train their language models.

What is fair use, and how does it apply here?

“Fair use” is a copyright law rule in the US — it lets people use copyrighted material without asking permission in certain cases (like teaching, or quoting it for research).

But, the new use must be limited and, usually, it must be "transformative." This means it adds something new with a different purpose — not just copying the original work.

In this case, the court said training AI on books can count as fair use. Why? Well, the court found the process was “transformative” — because the AI wasn't just copying the books. Instead, it was learning patterns and creating completely new content.

The fact that Anthropic bought the books legally also helped their case. They didn't steal or pirate the material.

How is UK law different from the US?

The UK doesn't have the same "fair use" rules as the US.

It has much stricter rules called "fair dealing." Fair dealing only allows using copyrighted material without permission in very specific cases (like research for non-commercial purposes, criticism, or news reporting).

So, as it stands, AI training on copyrighted books is not clearly allowed under UK law (even if the books were bought legally).

So, companies training AI in the UK face more legal risk than their US competitors. To avoid this, they may need to secure licences from copyright owners to ensure their use of any works is lawful (which can be time-consuming and expensive).

So, what’s the position in the UK?

The UK is dealing with a similar case right now: Getty Images v Stability AI.

The case began last month in the High Court in London — and it’s the UK's biggest case about whether training AI on copyrighted material breaks UK law.

Getty (a stock image company) claims that Stability AI (a generative AI company) scraped millions of its images to train its “Stable Diffusion” model. Getty says this breaks copyright rules, database rights, and trademark laws.

Stability AI argues that their data scraping:

• happened outside the UK (so UK copyright law doesn’t apply),

• were too small to matter (the AI doesn’t reproduce the original images exactly, so any copied elements are minor),

• fall under legal exception of fair dealing (since they’re not copying, they’re just imitating a style in a way that doesn’t harm Getty’s market).

The Getty case is a bit like the UK version of the Anthropic decision. But UK courts are likely to be much tougher on AI companies.

The UK's fair dealing rules are much stricter than the US fair use rules, which makes it harder for AI companies to defend their practices.

The law firms involved in Getty Images v Stability AI:

How can you use this in your applications?

This case shows a very real dilemma for UK-based AI companies — and it’s a great example to use in applications, or interviews to demonstrate how lawyers shape business decisions.

For instance, imagine a UK AI start-up planning to launch a new model trained on millions of books and images. They come to a law firm because they’re worried about being sued for copyright infringement.

The founders need to weigh up options like:

• Paying £££ for licences from publishers and image libraries to minimise risk.

• Moving their training operations to the US, where fair use makes things safer (even though relocating costs time and money).

• Removing risky datasets (e.g. from books), which could weaken their product compared to competitors.

• Or delaying launch entirely until UK courts or lawmakers clarify the rules — but they’ll risk falling behind to their American competitors.

These are not academic, legal questions — they’re hard, commercial trade-offs between cost, speed-to-market, and risk.

Understanding this will mean you understand how lawyers actually help clients frame and evaluate these decisions (which you need to show in your application).

They might negotiate licensing deals, advise on setting up a US entity, or assess which datasets are worth taking the risk.

More broadly, it’s a useful way to show that you understand a core part of commercial legal work: helping businesses make confident, informed choices when the law is uncertain — balancing innovation with compliance.

That mindset applies not just here, but in many areas of law where business and regulation collide.

IN OTHER NEWS 🗞

• 🤖 Freshfields is offering incoming trainees a fully funded master’s in law and technology at King’s College London (plus a £20,000 grant). Trainees will start their training contracts a year later, after finishing the course. The programme (which starts September 2025) includes topics like AI, blockchain, and big data. Up to half of the 2026 trainee intake is expected to join. This move follows Freshfields’ recent partnership with Google to roll out AI tools and shows its push to prepare lawyers for a tech-driven future.

• 💸 Monzo has been fined £21 million by the UK’s financial regulator for letting thousands of high-risk customers open accounts without proper checks. Between 2020 and 2022, Monzo approved accounts using obviously fake addresses (like Buckingham Palace and even its own office). The FCA said Monzo’s systems for spotting financial crime were not good enough as the bank was focused on growing rapidly. Monzo admitted the failings were from the past, saying it has since improved its controls and now has stronger safeguards.

• 🤝 Law firm mergers in the UK jumped 23% last year. The reasons were mainly tax changes, private equity (PE) interest, and succession issues at smaller firms. Smaller firms typically merged to avoid closure costs and to release ageing partners who don’t want to keep running things. Big firms and PE-backed “consolidators” bought firms to grow quickly, using scale and tech to cut costs. But merging isn’t easy — you need cultural fit and client trust to make it work. Here’s a good article from The Law Society Gazette.

AROUND THE WEB 🌐

• 🎾 Cute: Wimbledon is turning used tennis balls into cosy homes for mice as part of a creative wildlife conservation initiative

• 📹️ Useful: This website lets you search for unlisted YouTube videos

• 🪶 Funny: A seagull sips a man’s coffee then steals his mug (while he was installing anti-bird spikes)

STUFF THAT MIGHT HELP YOU 👌

• 📹️ Free application help: If you're applying to commercial law firms, check out my YouTube channel for actionable tips and an insight into the lifestyle of a commercial lawyer in London.