Competition in AI is spurring big players to think again about the role media and content can play in their success
No matter the cost of the pen, it still needs ink and something to write about. The next big AI payday could be playing right into media's hands.
In an age of apparent complexity, simple, truthful observations are the thinking material of choice. Just such an observation has been provided to us recently by Ezra Eeman, of the Dutch national broadcasting administration NPO.
During a conference at the JournalismAI Festival 2025 in London, where discussions understandably centred around the burgeoning use of AI in journalism, and Eeman himself talked about how his organisation is using AI in workflows, he turned the view on its head by stating: “The bigger play that’s happening, of course, is that media is being added to AI - becoming part of AI systems.”
Hmm. This is one of those situations where seeing the same thing from different perspectives is truly valuable and can help recolour a picture entirely. Frankly, it’s good to hear a senior media figure reiterate what many have hoped since the arrival of AI: in competing with each other, those with the data - and the deals - become king.
From Ask Jeeves to ask anything
There is no doubt now that we’re entering the era of the Answer Engine, or Query Engine, or simply AI search. Personally, I hope Query Engine wins out, but it’s not the most promising phrases that necessarily win the common description game, as we’ve found travelling along the information superhighway.
Indeed, as we enter this new phase of information retrieval, for the internet archaeologists among us, it’s worth reflecting that the marketing concept of Ask Jeeves is closer to where we have arrived now than the OG Q&A product was before it was out-muscled by Google Search. Despite not having the technology to meet the promise of natural language answers to natural language queries, the aspiration was there, if in name only.
Here is where we are though, with natural language queries being answered in natural language responses. There’s no putting that genie back in the can, nor should there be a desire to.
What is apparent though is Mr Eeman’s point: media is being added to AI, not the other way around. Much as video didn’t actually kill the radio star, conventional search will continue to have utility, yet the ease of use of transformer-based systems and way they are packaged into what look like regular search tools and everyday apps, will likely see them triumph in day-to-day information retrieval. They are simply an easier tool to use than going to Google and following threads, no matter the quality of the silk.
Despite the awful, indefensible situation we have seen where the work of others is taken without consent and spun into value for businesses that care not for the originators of the data used, a different world is now becoming visible, and it looks somewhat inevitable. If, as Eeman believes, media is to become AI, then the media that powers the AI still has a distinctive role to play, one in which it can regain value.
By way of useful example, Meta, currently something of an also-ran in the public-facing AI stakes, still harbours serious ambition in the sector, backed by Meta’s cash mountain and a focus which saw the company recently ditch its dubious-from-the-outset Metaverse to clear the decks and double down on AI products that people actually want.
As part of this new AI strategy, Meta announced last week a raft of content licensing deals with USA Today, People Inc, CNN, Fox News, The Daily Caller, Washington Examiner and Le Monde. More are apparently planned.
It seems like this could be the start of a natural progression from the “fill your boots with scraped data and tell everyone it’s not a crime, it’s the future” situation we’ve been in hitherto. Meta’s plan is to make a virtue of the sources it is using, and also a loudly proclaimed marketing hook: “Our sources are good and disparate, that’s why you should trust our system more than the others!”.
Meta said: “When you ask Meta AI news-related questions, you will now receive information and links that draw from more diverse content sources to help you discover timely and relevant content tailored to your interests.”
It also looks like Meta is seriously considering a subscription model for full access to this service, apparently code-named Avocado. If making use of good and constantly updated sources can be sold as a desirable feature, then it’s quite possible it will succeed.
If it does, it will reinforce that there is a market to be won - the market for fresh content and exclusivity deals - and a market for cooperation with rights-holders and creators.
Meanwhile yesterday Google also announced a payment scheme to reward those whose work is used in AI search results.
It’s not a deluge of deals, and there are big questions for smaller publishers who don’t have the clout or voice to kick up a noisy fuss. Nor can we excuse how media makers and creators have been roundly ripped off to date, and there is years of litigation and legislation to be played out - slowly - before we feel “made good”, if we ever do.
But, in a world where entire publisher sector revenues and valuations look like Big Tech budget rounding errors, it makes sense to consider how we can be rounded up and rewarded, rather than rounded down and diminished. Those golden pens still need ink and ideas. Today it is the big players being rewarded, tomorrow we can hope it is everyone else.
We, the media, are becoming the AI, and becoming the market. We always were, really.
How to turn clicks into cash
Sweden’s Aftonbladet has found a clever way to go from anonymous browsers to paying subscribers. With a little help of a machine learning model which they integrated into their front-page system, the publisher can now predict which articles are most likely to convert, and all that without having users logging in. As for the result, a 75 per cent bump in sales from front-page stories. With a little help of reliable signals and a good in-house infrastructure, Aftonbladet has proven that even limited data can drive big engagement and happier readers.
Read
Age checks gone awry
While born from sincere intentions, the execution of the UK’s Online Safety Act has already stirred up controversy and many cybersecurity, privacy, and free expression issues. Age checks are clunky, the wrong posts are being censored and many smaller sites are shutting down in fear of compliance rules. The Open Rights Group, as well as many other critics, are urging British MPs to reconsider and come up with a smarter and rights-respecting approach which would not only protect children online, but everyone else.
Read
Pumping up the prompt volume
Ever wondered if AI trackers are as smart as they claim, or if they’re just really good at scraping messy data from Chrome extensions and then mixing the searches up with emails, trip plans, and a little bit of connecting the dots? Steve Toth has the answer for you, as well as the results, which may or may not be wildly inflated numbers compared to Google or Ahrefs.
Read
Making peace with the Mouse
Disney is to make a $1 billion equity investment in OpenAI and has also inked a content licensing deal that will see OpenAI’s Sora genAI video app able to “generate short, user-prompted social videos that can be viewed and shared by fans, drawing from a set of more than 200 animated, masked and creature characters from Disney, Marvel, Pixar and Star Wars”. After being at legal daggers drawn, the two businesses have clearly now seen commercial sense in co-operation.
Read
Paywalls vs big tech
With big tech hoarding traffic, repurposing content and leaving nothing more than scraps for creators, some measures are due. Paywalls are one of the options, however if you’re looking to get one done, you should do it the right way. Harry Clarkson-Bennett has got your back.
Read
Google’s monopoly gets a timeout
A federal judge has given Google a bit of a reality check. After ruling that the tech giant illegally held a search monopoly, Judge Amit Mehta is now hitting Google with a one-year limit on default search and AI contracts. Basically that means that Google will have to renegotiate every year, in order to give competitors an actual fighting chance to get their foot in the search door. But that’s not it, there will be some sharing of search data with rivals as well. The trouble doesn’t stop there, as on the other side of The Atlantic, The European Commission is snooping around as well, looking for any shenanigans which could point to a breach of antitrust rules.
Read
X gets burned by EU rules
Elon Musk’s X just got hit by a $140 million fine, being crowned the first platform to feel the EU’s Digital Services Act bite. The charges? Selling blue checks like candy, hiding public data, and hiding its ad info so well not even Sherlock could find it. This has in turn exposed users to all kinds of internet tomfoolery, from scams to bots. X has 60 to 90 days to clean up its act and make sure it follows the rules, but as you can expect Musk is already calling foul and labelling this as EU overreach. Brussels insists it’s not censorship, it’s just rules. You decide.
Read
Is this the board that keeps Meta honest?
Born out of scandals, hate speech, and privacy blow-ups, Meta’s Oversight Board, operating since 2020, was the company’s solution to repair its reputation. It’s a unique body in tech, but does it really have power? “Five years on, the Board has made important strides for Meta’s global users, bringing transparency, reasoning and a human rights perspective to decisions that were long made behind closed doors, and with little or no public-facing rationale” its latest report claims.
Read
Search console gets social-savvy
Rare W for Google, as they’ve decided to make life a bit easier for site owners. Now the Search Console Insights report is watching both the website and social channels, stitching together web and social data and eliminating juggling between dashboards as well as the guessing game of which post did what. While early testers say it’s neat and convenient, we’ll have to wait and see what its like in the long run.
Read
Taming AI agents
OpenAI, Anthropic, and Block just launched the Agentic AI Foundation (AAIF) under the watchful eye of the Linux Foundation, with the goal to make their beloved AI agents play nice together. Each of the companies pitched in: OpenAI with their AGENTS.md, Anthropic its Model Content Protocol, and Block tossed in Goose. The ideal scenario would be a well-behaved and interoperable sandbox filled with sunshine and gigabytes, without any red flags such as AI browsers spilling secrets or going rogue (Hint: that’s exactly what happened). Microsoft, AWS, and Cloudflare are also joining the party.
Read
UK life online
The UK regulatory body Ofcom has come out with the annual Online Nations Report, revealing how adults and kids are spending their time online. Obviously, with AI in the mix things are changing, adults are becoming more sceptical while kids are mostly happy. Read the full report here [PDF].
Read