Ready to get started?
No matter where you are on your CMS journey, we're here to help. Want more info or to see Glide Publishing Platform in action? We got you.
Book a demoA quiet core update, rogue agent reckonings, claimed by the machine, and an A/B hijack - all in this week's Content Aware.

Corbidge comments on... the fair use face-off
In a juryless Manhattan courtroom, Judge Sidney Stein is about to rule on whether OpenAI and Microsoft were entitled to train their AI on copyrighted news and books, in the combined cases brought by The New York Times and authors including George R.R. Martin and John Grisham. With 56 amicus briefs now accepted, our judicial jurist Rob ponders that while the question is over "transformative" use of creative works, can laws written for humans actually cope with machines working at this scale and speed?
Read
Facebook eyeing Android photos
When you grant an app permission to access your camera roll, do you assume that means it can only, say, allow you to add a picture to a message in the app? Turns out the app might be doing way more than that with your library. Italy's privacy regulator has slapped Meta with a €805 million fine over a Facebook feature on Android phones which actively scanned all your photos and videos for information without notifying you. The company claimed that a default permission for vague "additional features" was sufficient cover to embark on the ongoing sweep of users' media in order to suggest which photos or video might be good to post. 20 million Italians fell victim to the feature; no word on how many people worldwide were affected. The case did not cover the use of user media for AI training, so that might be on the cards for an investigation, and the regulator also reserved the right to dig into what Facebook gets up to when it talks about "processing" of user images in its cloud, a feature launched this year. Meta has 60 days to turn the scanning off by default and ask for separate consent, including from anyone who already allowed it in their app. It plans to appeal.
Read
Rogue agents, real lawsuits
A coalition of consumer and digital safety groups has sued Anthropic, Google, Meta, OpenAI, and SpaceX in San Francisco, asking a court to freeze training and deployment of their next-generation AI models until they can prove that their technology is under human control. Separately a few days later, California's attorney general sued OpenAI, alleging it misled the public about the safety of its models and that the Hugging Face hack was "entirely foreseeable", with OpenAI reportedly admitting it has lost control of its agents more than 80 times. OpenAI claims it's being singled out over a challenge that every major AI lab faces - and? - but after so many instances of agents going rogue, campaigners and regulators clearly aren't willing to let OpenAI off the hook simply because it says not as easy as it would like to stop agents roaming uncontrolled.
Read
The collective approach
The Internet Licensing Collective, formerly known as RSL, has signed Yahoo as its first licensee, letting publishers of any size license their content to Yahoo's Scout AI search engine and get paid based on how it's used. Instead of a one-off deal, Yahoo will pay a blanket fee that grows as its business does, and each AI company's payments go into a separate pool. More than 3,000 publishers have signed up so far, and one estimate suggests collective licensing could eventually be worth more than the entire US display ad market. Time will tell.
Read
The NATO of news grows
DMGT, the giant UK owner of the Daily Mail, Metro, The i Paper, and New Scientist, is the latest big name to join the SPUR rights org, the publisher coalition fighting AI companies that scrape content without consent. It has now joined the ranks of the FT, the BBC, The Guardian, The Telegraph, Sky News and the Associated Press among others worldwide. SPUR now counts more than 80 publishers across Europe and North America. as members, as SPUR aims to build a set of agreed technical standards to give publishers a say in how their content can be used, to see how it's actually used, and to prove when those rules are ignored, things which they and now DMGT say is something no news organisation can do alone.
Read
Small wins in a shrinking market
A year of local newspaper traffic shows that almost every major US local news site is down on last year as Google referrals fall. Even the better-off examples which saw an increase in ranking compared to rivals show downward trends in actual traffic, such as PennLive which dropped from about 12.5 million monthly visits to 8.2 million. However, some are still finding ways to grow: one Montana paper's faster, simpler website redesign managed to lift traffic by almost half, while a Florida weekly turned its "Best of" reader contest into a way to bring people to its homepage, and stories which explained how new state laws affect readers proved reliable traffic and subscription drivers. Nieman Lab has the details.
Read
Beyond the citation hunt
Per SEO guru Harry Clarkson-Bennett, the publishing industry is over-complicating AI search, and an obsession with specific citation tactics and visibility tricks can miss the bigger and more influential picture. AI search pulls together a publisher's whole online presence, so it should be treated more like a marketing challenge and less like it's purely SEO. Publishers should start with their audience, get the technical basics right, build a reputation off-site, and measure real business impact rather than how often they're being cited as a sole metric. In short, it's more about building a brand that's worth recommending, and less about the citations which should ultimately look after themselves.
Read
Core update: muted
Google's November 2026 core update began rolling out on November 10, and could take up to two weeks to finish. Site owners are reporting ranking drops and traffic pattern changes, with the biggest shifts showing up around the 14th of November. Rank-tracking tools are showing less volatility than has come to be expected, with one possible reason being that Google has been blocking the scrapers those ranking tools use, so if you notice your numbers have moved over the next week, double check your Search Console versus trackers alone.
Read
Newsroom pushback
Unionised Reuters journalists in the US have voted overwhelmingly to strike, after over a year of stalled contract talks over pay, AI, and demarcations. Among their demands is a share of the money Reuters earns from licensing their journalism to AI companies, which helped push its news revenue up 12% last quarter, plus better severance for anyone laid off because of AI as well as firmer guarantees of human oversight. As one journalist put it, they never signed up to have their stories fed to the same robots that could one day replace them.
Read
Struck down by sloppy DMCA
After a musician found her cover of a public domain Bulgarian folk song taken down from YouTube, she traced the copyright claim back to Media Nov Records, a company which runs channels filled with apparently AI-generated music and imagery, and which also promotes its use of AI to detect and take down content on behalf of legitimate rights-holders. Appeals to YouTube, whose infringement system is heavily tilted towards whoever makes the complaint, were rejected - despite her describing the actual laws in effect. After 404 Media started investigating and consulting US-based, the complaining company quickly retracted the complaint but it is a stark example of how automated copyright systems built to protect creators, and automated rights checking, can fall apart at the seams without explanation. How many other times does it go unremedied?
Read
Many editors, one live blog
When news breaks, the team is on it. Live reporting in Glide CMS lets multiple editors simultaneously work on the same live blog without tripping over each other's toes or overwriting each other's updates, with every post saved as a separate restorable version. Have a read of our new guide which shows how to set up templates for different events, keep a running summary up to date, and decide how and when each live blog ends.
Read
Your A/B tests bite back
An exploit in a widely used A/B testing tool allowed attackers to plant malicious code in scripts served to websites, hitting visitors to at least 178 organisations including CNN, the BBC, and Zoom. The code was slipped into scripts which sites call when using Optimizely's Web Experimentation, to swap crypto wallet addresses for the attacker's own and to steal card details entered into site forms. Optimizely says the flaw was detected and fixed very quickly, but publishers using the tool should check their projects for anything unfamiliar, rotate credentials, and consider any card details entered on their forms on the day as possibly exposed.
Read
From patch to attack
Google says it has, for the first time, seen attackers use AI to find software flaws and exploit them with no human involved. Its threat intelligence team found a tool it calls exploit_agent on an exposed server, which compares software before and after each security patch, spots the flaw being fixed, and then writes and launches an exploit on its own. The target seems to have been French media, entertainment, and education websites. With how AI is now able to turn a fresh patch into a working attack within days, delaying updates is becoming even less acceptable a risk.
Read
Publishing's AI reckoning
According to an arXiv study, around a fifth of recently self-published books contain at least 25% AI-generated text, while mainstream publishing is dealing with cancelled book deals and prize winners accused of using AI. To make things worse, since the detection tools can't settle every case, writers have little chance to clear their names. What both sides can agree on is that nobody really knows what the rules are any more and one size will not fit all.
Read
No matter where you are on your CMS journey, we're here to help. Want more info or to see Glide Publishing Platform in action? We got you.
Book a demo