AI Overviews busted with its hands in the till - again

#327: Google AI exec admits his search team steals publisher content despite half the world's news providers officially opting out...

Google is training AI Overviews on news content taken without permission despite publishers opting out of training, an exec just admitted in court.

Publishers have always been told that they can opt out of training by making a few simple code tweaks. Google has gone out of its way to emphasise how easy it is.

But the court hearing that’s seeking to break Google up has been told that opting out only works on DeepMind - and not on search - which is the main course.

Now a Google VP has confessed under cross examination by Department of Justice lawyer Diana Aguilar that the search team still uses the opted out publisher content to create AI Overviews.

And Aguilar got him good…

Welcome to new subs overnight from Netflix, NBC News, AOL, DuckDuckGo, Human Native AI in London, Bauer Media in the UK, and many more.

And, join me and superstar designers Canva and Flourish for a free three-hour session to explore the future of visual storytelling in New York later this month.

Let’s get back to the latest evidence of Google’s $2 trillion AI theft…