BBC Threatens Legal Action Against Perplexity AI Over Content Scraping 21
Ancient Slashdot reader Alain Williams shares a report from The Guardian: The BBC is threatening legal action against Perplexity AI, in the corporation's first move to protect its content from being scraped without permission to build artificial intelligence technology. The corporation has sent a letter to Aravind Srinivas, the chief executive of the San Francisco-based startup, saying it has gathered evidence that Perplexity's model was "trained using BBC content." The letter, first reported by the Financial Times, threatens an injunction against Perplexity unless it stops scraping all BBC content to train its AI models, and deletes any copies of the broadcaster's material it holds unless it provides "a proposal for financial compensation."
The legal threat comes weeks after Tim Davie, the director general of the BBC, and the boss of Sky both criticised proposals being considered by the government that could let tech companies use copyright-protected work without permission. "If we currently drift in the way we are doing now we will be in crisis," Davie said, speaking at the Enders conference. "We need to make quick decisions now around areas like ... protection of IP. We need to protect our national intellectual property, that is where the value is. What do I need? IP protection; come on, let's get on with it." "Perplexity's tool [which allows users to choose between different AI models] directly competes with the BBC's own services, circumventing the need for users to access those services," the corporation said.
Perplexity told the FT that the BBC's claims were "manipulative and opportunistic" and that it had a "fundamental misunderstanding of technology, the internet and intellectual property law."
The legal threat comes weeks after Tim Davie, the director general of the BBC, and the boss of Sky both criticised proposals being considered by the government that could let tech companies use copyright-protected work without permission. "If we currently drift in the way we are doing now we will be in crisis," Davie said, speaking at the Enders conference. "We need to make quick decisions now around areas like ... protection of IP. We need to protect our national intellectual property, that is where the value is. What do I need? IP protection; come on, let's get on with it." "Perplexity's tool [which allows users to choose between different AI models] directly competes with the BBC's own services, circumventing the need for users to access those services," the corporation said.
Perplexity told the FT that the BBC's claims were "manipulative and opportunistic" and that it had a "fundamental misunderstanding of technology, the internet and intellectual property law."
If the shoe fits... (Score:3)
Everyone's IP is up for grabs, until everyone starts to grab Perplexity's IP, in the eyes of Perplexity. Sure, they can hide behind "not understanding IP law", or "not understanding internet" or whatever. What Perplexity and all other LLM providers do not grasp is "content ain't free, so either pay up or shut up".
You would think that people smart enough to build and/or deploy LLMs would have no trouble to understand that basic truth. Of course, that is not what they promised to their shareholders, who want to see returns quickly for the ungodly amounts of money the LLM builders/deployers usurp from their shareholders/investors and/or their own greed.
Because all of these LLM builders have shown, when the rubber hits the road, that they are as greedy as any psychopath dares to dream of being. And more.
And have no qualms of using the argument "But China will...". In a similar fashion as the police/politicians overuse "Against terrorism" and "Think of the children...". It is getting a bit long in the tooth. These companies are absolutely unwilling to play by anyone's book but their own, while copying all their homework from your homework, and then claim you copied their work instead.
Articles and comment like the ones mentioned here being uttered by Perplexity personnel, they give me the impression that they are exactly like petulant children, who are flabbergasted about receiving demands for (fair) compensation by the people/organizations they steal their reason for existing on.
All of the commercial LLM providers can (and should) be accused of the same behavior, as Perplexity sure isn't the only one.
Also, they didn't take the hint they got from DeepSeek R1 either. Be more efficient, instead of demanding more power (energy) and liberty (copyright) for their data hunger. Efficiency, that never has been the forte, of U.S. companies, only "more, more, more". LLM providers from the U.S. can't help themselves, this is too ingrained into their DNA, I suppose.
And yet, if they would be inclined to go the way of efficiency, then all the money they would save on hardware and the energy this requires could then be used to (fairly) compensate those making the content/"AI feed". That would be a win-win for everyone involved. Probably too simplistic of an idea for these LLM providers, especially since no AI/LLM was used to think of it, just a tiny lick of common sense.
Re: (Score:2)
What the hell is "17.1.106"? If you mean section 106 of Chapter 17 of the U.S. Code, that's usually written as something like "17 USC section 106".
AI companies usually infringe on the first of the rights listed in 17 USC 106: "reproduc[ing] the copyrighted work in copies". This is done repeatedly during training the AI, when the model creator copies the training material to nonvolatile storage for reference, then again when they load it into RAM to train the AI model. In some cases, the AI model will gen
Re: If the shoe fits... (Score:2)
" In some cases, the AI model will generate further copies by how it responds to prompts, "
In no cases, you mean. It creates some close variants which might be considered derivative works, though.
Re: (Score:2)
Sorry but no, both you and the bbc need to realize that the USA is not the UK. This is a privately owned American company that does not do any business in the UK.
We are held to US law, not UK law or whatever made up law you're wanting.
If what they are doing is also illegal in the USA then they can be extradicted and be prosecuted. If international law worked on the principle of "well we did it from our country, not in yours" then nobody anywhere would ever be able to be prosecuted for financial fraud, theft etc. Fortunately for people like you it is a founding tenet of international law that you cannot direct harm in another country from your own and escape justice.
Re: (Score:2)
Re: (Score:2)
What Perplexity and all other LLM providers do not grasp is "content ain't free, so either pay up or shut up".
It's a good reminder though, that copyright is an artificial construct created by governments.
It's been around so long, and grown to such monstrous proportions, that we tend to think of it as a natural, real thing, But it's really more like tax policy than (say) the strictures against murder.
There's nothing intrinsically wrong with the UK government contemplating adjusting it in the public interest. UK subjects do benefit from using tools like Perplexity, you know.
Re: (Score:2)
I guess. (Score:4, Funny)
They should have paid their TV license eh?
Hahah!
You are The AI. Exterminate! (Score:3)
I thought those folks at the BBC loved regeneration?
A.I leaching off other peoples work (Score:2)
Set it BBC as a service (Score:2)
Meta's AI memorized 42% of first Harry Potter book (Score:2)
“A new study reveals that Meta's latest artificial intelligence model can reproduce nearly half of the first Harry Potter book from memory, raising fresh concerns about copyright infringement in AI training as the company faces mounting legal pressure from authors and publishers.”
Re: (Score:2)
Meta's AI memorized 42% of first Harry Potter book [perplexity.ai]
“A new study reveals that Meta's latest artificial intelligence model can reproduce nearly half of the first Harry Potter book from memory, raising fresh concerns about copyright infringement in AI training as the company faces mounting legal pressure from authors and publishers.”
So ... the argument here is that somebody isn't going to buy the book - or check it out of the library - because Grok might be induced to cough up 42% of it? (Presumably in suspect chunks - how would the user know whether they are accurate anyway?)
Seriously?
Some cheek (Score:2)
Lol misunderstanding of the internet, BBC.co.uk was an early domain, surviving many a ddos.
Think the BBC has a very large domain portfolio, so many dramas mention site names and they have to register it before someone puts a goatse on it.
BBC peer with a lot of ISPs, they know the internet pretty damn well.
Re: (Score:2)
The BBC is wrong (Score:2)
Perplexity uses a few external models and a model that is based on R1. They do not train on BBC content, but they process fetched data to create summaries. That's covered by copyright law (and possibly free speech laws) as long as they do not provide originals verbatim.
Training AI vs training humans (Score:1)
Re: (Score:2)
humans ahave different rules under the law