Aardvark Daily

Please visit the sponsor!

A new challenge for copyright

When I created 7am News back in the late 1990s, I published news stories from around the world.

Although I had a couple of people overseas who regularly filed stories, much of what appeared on the pages of 7am.com came from my own keyboard and consisted of stuff that represented a summary of facts gleened from other news publications.

Was I breaching the copyright of those other news sites from which I obtained those facts?

Absolutely not.

You can't copyright facts but you can copyright the prose used to express those facts, and not one single sentence was republished verbatim from those news sources unless it was as a direct quote.

I was never challenged for doing this because the news publishers of the day knew full well it was totally legal and in many cases, something they did themselves.

However, the arrival of AI is changing perspectives within the world of copyright.

It seems that while it may be okay for a human being to extract the facts from a copyrighted story and then write their own prose to report those very same facts, having a computer and AI do this "changes everything".

I spied this story on Variety.com this morning and it hints at the problems AI may be about to dump on those who are not prepared for the effects it will have on the value of their intellectual property.

Regular readers will recall that I have already predicted that the likes of Google will use their AI systems to scrape news sources from across the planet and then write highly personalised news summaries for users. This would come in the form of a news page that was highly focused on the subjects that you are interested in or which are most relevant to you.

The stories would not be verbatim copies of material from other news sites but rather carefully crafted prose that contained the facts in a totally different stream of language.

Just as when I did this manually back in 1999, AI will do it automatically today.

The lawsuit reported on Variety seems to suggest that publishers such as Down Jones, The New York Post and others, now consider the facts in their stories to be copyrightable -- which, under current laws, they are not.

Unless Perplexity (the target of the lawsuit) has been republishing material from those publishers absolutely verbatim then should a court ruled in favour of the plantiffs, it would set a very worrying precedent and seem to be at odds with the protections that copyright legislation is supposed to provide.

Of course it's easy to understand why news publishers would be highly pissed-off that some startup was scraping their pages and making money from the facts they were printing -- but right now, that simply does not appear to be illegal under existing copyright law.

However, the arrival of AI-based systems such as the one operated by Perplexity (and pretty soon Google) may force legislators to update their copyright laws to make such scraping illegal without a license from the IP owner.

Even if the laws are updated, there still remains a burden of proof on the part of the plaintiff that could be hard to reach. If an AI website publishes a story that includes facts that were previously reported by another news publisher, how do you prove that the AI system obtained those facts by scraping that other website?

Things are looking a bit messy right now and with Google having a hand in the game, it will be incredibly interesting to see how this all pans out. We know that many Western nations have the best justice and political systems money can buy -- and Google has a lot of money.

Carpe Diem folks!

Please visit the sponsor!

Here is a PERMANENT link to this column

Rank This Aardvark Page

Change Font