The great AI content heist

Boost for Google but blow to the web, UK and OpenAI's partnership, AI's bubble test, and X vs France- all in this week's Content Aware.

Published: 15:33, 24 July 2025

Corbidge comments on...common sense and copyright
In a speech given at the AI Summit in Washington DC on Wednesday, we learned that President Trump sees turbocharging the AI industry as a key plank of future American economic success, yet the White House incumbent also unexpectedly added his own thoughts on copyright and AI training data. Based on that, some are saying the tech titans have got to him, yet it's equally possible he has got to them. Our resident mystic Rob ponders on copyright, IP, and common sense.
Read

The new content outlaws
The great AI content heist is underway, and it's not only bots, now we've got third-party scrapers coming into the ring. These digital burglars are slipping past paywalls and taking publisher content - sometimes to order - without so much as "how do you do?". Even though Cloudflare and allies are trying to slam the door shut, this is already a nightmare for publishers: losing traffic and revenue while AI models serve their work for free. There are more and more calls for transparency and tighter rules, before the internet becomes the Wild West.
Read

AI Overviews: win for Google, loss for websites
Google's AI Overviews are chomping down on clicks, and the websites aren't loving it. A fresh Pew Research Center study revealed that users are about half as likely to click through when AI Summaries pop up, with just 1% visiting the sources. To add fuel to the fire, many don't even continue searching after reading the AI's answer, possibly swallowing some of those infamous hallucinations. At around 20% of all search results, AI summaries are great for Google's wallet, but not so great for publishers' traffic.
Read

UK and OpenAI: lotta promises, little plans
OpenAI just shook hands with the UK government in a non-binding deal that appears a bit big on ambition, but light on obligation. It is framed as a strategic AI partnership, but it's more of a promise that something might be done someday, such as data centres, help with public sector AI, and a chat about model safety. It fits neatly into what OpenAI seems to be doing globally: cosying up to regulators, looking helpful, and quietly paving their way for a future where government tools might just run on ChatGPT. It's a smart move, no strings attached - but plenty of lucrative opportunities.
Read

Google's summer update
One more algorithm update to unpack, as Google has quietly rolled out a core algorithm change while we were enjoying our holidays. Some news sites were slapped, some were not. SEO sage Barry Adams shares his thoughts.
Read

Fake cases, real consequences
AI is now showing up in US courtrooms, and not in a good and useful way. Lawyers have started slipping AI-generated filings into the system, complete with fake case citations confidently hallucinated by tools like ChatGPT. In one Georgia divorce case, a judge signed off on such a document, which was later exposed on appeal. The lawyer did get hit with a $2,500 fine, but the real damage is to the credibility of the court. No one believes this is an isolated case.
Read

The dawn of the post-journalism era
The closure of Al Jazeera Balkans is a heavy blow to press freedom in the Western Balkans, and it might be just the tip of a global media meltdown. Once upon a time, legacy outlets were propped up by international cash aimed at promoting democracy, but now they are in danger from tightening authoritarian grips, funds drying up and government interference. Red flags are already being waved by experts, warning about a "post-journalism" era, where disinformation runs rampant while the truth cowers in the shadows. Unless things change, journalism risks becoming just background noise, both in Europe and in the rest of the world.
Read

X vs France: free speech or foul play
There seems to be another legal brawl brewing, this time between Elon Musk and the French prosecutors. X is pushing back hard against a French criminal probe, claiming it is nothing more than a politically charged attack on free speech. Prosecutors are looking for access to X's algorithm and real-time data, citing alleged bias and data abuse, and they're using organised crime laws to get it, complete with wiretap powers. X says "hard pass" on all that, accusing the investigators of being openly hostile to the platform. With Europe cracking down via the Digital Services Act, this clash could be one of many such.
Read

AI bubble's biggest nemesis
Ever wanted a comprehensive guide on how to hate the AI bubble, filled with arguments on why this is not the future? Ed Zitron's got the best flying circus take in the game.
Read

Cutting public media cuts democracy?
Does publicly funded media have a place in modern democracies or is our age of media multiplicity one that doesn't need them? The Republican administration has taken a hard stance against editorial positions held by both NPR and PBS in the US and consequently Congress has voted to rescind $1.1 billion from the organisation that runs them, effectively defunding both. Dr Sam Martin makes the case for the preservation of such institutions.
Read

AI meets the bubble test
AI innovation isn't a bad thing, but are we approaching a bubble territory? Analysts seem to think so. With tech giants such as Nvidia and Microsoft sporting sky-high valuations, some say this feels like the dot-com era all over again. Every other day we're hearing how Big Tech is pouring billions into AI, but most startups likely will not survive a shakeout. If the bubble bursts, we're all in for a painful reset, but maybe also some grounded progress.
Read

Duck the slop, keep the privacy
If you're tired of AI slop cluttering your image searches, DuckDuckGo has got your back as they've just rolled out a filter to hide AI-generated images. The filter runs on open-source blocklists, so it might not catch everything but it should thin out the deepfake herd.
Read

Google's secret sauce
Ever wondered what is actually going on inside Google Search? Nearly 1,200 experiments are shaping how we find what we want. After looking at the intricate structure and brain behind the info and real-time updates, it's clear that it might not all be about pumping out endless content, but figuring out how to become a verified, trusted player in Google's web.
Read