It has been a while. Hello again. I’m back talking about copyright. Can’t shake my geeky obsession. But why now?
The specific thing which has got my goat is a proposal from the UK government to take a wrecking ball to what is left of copyright law by largely exempting AI companies from it.
It looks crazy at a glance, and only gets crazier if you dig into the detail. Unsurprisingly, the UK creative industries, which depend on copyright and which are worth £125bn to the UK economy every year, are implacably opposed. In fact, it’s quite hard to find anyone at all, outside the government, who thinks it’s a good idea.
Ministers are finding this out for themselves, because they have started a consultation about their plans. It suggests a range of options, but it also says that the government has already decided which one they’re going to implement. So while the consultation responses might highlight just how much people dislike the proposal, it seems the government has pre-emptively decided to ignore them. The allure of imagined AI riches is just too strong.
I’ll highlight some of the choicest morsels over some future posts, to help explain my own views about it and maybe inspire a few people to submit their own views before the deadline of 25th February 2025.
For now, a quick summary:
AI companies “train” their systems by copying everything they can find on the internet and feeding it into their computers. This is how those AI systems “learn”.
Copyright is, literally, the right to make copies. It’s a kind of property — intellectual property — and, like other kinds of property, it belongs to someone. Not AI companies and not the government. Someone who doesn’t own copyright doesn’t have the right to make copies unless the owner – or the law – has given it to them.
Rather than ask permission, though, AI companies have simply ignored copyright and copied everything anyway. Without all that content, their systems wouldn’t work. In fact, the content they have used, far more than the computer chips or the power sources and arguably the underlying technology, is the most valuable component of what they do. They want it, they need it, it’s right there on the internet for anyone to see. So they have simply helped themselves.
To add injury to insult, they are using their systems to obviate the need for people to seek out the source of the “knowledge” they’re imparting to their users. They’re competing against their unwilling and unrewarded suppliers and damaging them commercially.
This isn’t a popular move with the people whose content they have illegally used. However, it has got the UK government very excited. AI has been hyperbolically projected to generate gigantic riches. The new-ish government, desperate for anything which might help them create growth in the UK economy, wants some of those AI riches to come the UK’s way.
So they’re proposing to wave a magic wand and make the illegal copying that AI companies do legal, by creating a special exception in copyright law for them.
This won’t end well. I’ve re-started blogging about this to explain why, and to suggest better ways. Stay tuned…
Copyright Blog 