The fragile state of trust, how people use AI Mode, and a publisher roadmap for SEO and AI

Big tech trials tribulations, Reddit rebels against scraping, and local news fighting for its life - all in this week's Content Aware.

Published: 15:14, 09 October 2025

Corbidge comments on... a heroic U-turn
Did anyone think OpenAI's Sora mass infringement machine was launched in complete innocence of its ability to inspire every Hollywood copyright and IP attorney to order a bigger than usual mojito and rub their hands in glee at likely payouts? In fact, that was probably part of the plan, says our Best Boy Rob who has a grip on such things. While Hollywood has inspired a rare volte-face from Sam Altman et al, we are all benefiting from seeing how it can be done.
Read

AI, SEO, and Publishing: A roadmap 
AI is shaking up publishing as much as the arrival of the internet once did, and we all know the new normal is to be uncertain of anything any more. If Google and traditional search done for? Hardly - they are still the biggest factors in play for getting content seen. But that doesn't mean the shape of the landscape won't change. SEO leading light Barry Adams dedicates his latest newsletter to looking at the impact Big AI is having on search, media, and publishing businesses, sounding a series of positive notes to those who are convinced the only tune is of doom and gloom.
Read

Harry's game
Speaking of SEO gurus with newsletters and the subject of AI, keep an eye on Telegraph SEO supremo Harry Clarkson-Bennett's project to explore the sort of next generation SEO tools destined to become default in newsrooms by building his own. It's a combination of AI, coding assistant, analytics platforms and elbow grease, trying to answer the media industry question on how to crack search visibility. Clarkson-Bennett was inviting helpful eyes to give useful feedback, but even if you can't join the project, look for updates and learnings down the line.
Read

Audience trust is slipping
A new Gallup poll outlines just how much US media brands are up against it, with indications that trust in the traditional news outlets of newspapers, TV, and radio has dropped to historical lows across all political viewpoints. There are splits between party supporters but the trajectory is shared regardless of voting intentions. Just as clear is the slide of trust across age groups, showing that a major challenge in attracting younger audiences isn't just the tone and delivery of news - being where they are and talking their language - but also that younger people tend to trust traditional media the least, a compounding challenge.
Read

AI Mode - how people actually use it
A fresh usability study of Google's AI Mode offers nuggets of insight into changing user behaviour, and reveals that most users now barely venture outside Google's environment - which was always the Google plan right?
Read

Supreme Court adds fuel to the fire
The US Supreme Court just handed down a potential game-changer in the app store saga which could reshape how apps get on to phones and how people pay for things - including content. It upholded a judge's order to Google to allow rivals into the Play Store and to use non-Google payment systems. Suit initiators Epic Games is cheering, while Google is waving red flags of security and brand damage. The full judgement is expected by 2026.
Read

The SB 53 aftermath
California's recent SB 53 AI trust and safety bill requires AI companies to show how they are keeping models safe from going rogue. Some AI figures made out like it would cause the end of the technology and cat videos, but Adam Billen from advocacy group Encode AI believes it's a good sign that regulation doesn't have to be the villain in the story of AI, and that lawmakers and tech people can agree on good things. As with everything nowadays, there are some non-believers. In DC, the SANDBOX Act is being pushed, which could let companies dodge state rules for a decade.
Read

AI browsers and the security illusion
A scare for users of Perplexity's Comet AI browser this week, after a security report claimed "CometJacking" by insertion of a doctored URL can hijack the AI assistant and steal private data. Perplexity downplayed it and reportedly issued a patch, but it's a showcase that as AI firms and other sources try to disrupt user behaviour by launching their own browsers, opportunities will arise for ne'er do wells. However good their AI may be (!), many of these companies are still barely out of shorts and it is completely appropriate to question how or if they are covering the basics. Are your staff using such tools?
Read

Follow the Google ad trial
The Google ad tech trial trundles on, potentially much more impactful for publishers and anyone spending money on ads than the damp squib Google search trial. Check in to summaries of each day's courtroom events at the Big Tech on Trial site, covering proceedings with a good overview of each side's position and - importantly for any court reports - conveying that vibe of how it actually feels in the court room and the judge's demeanour.
Read

AI in the newsroom: handle with care
As expected, newsrooms are cautiously letting AI handle the boring stuff, such as transcripts, summaries, and SEO, while keeping a firm grip on the journalism. Outlets such as The Goldendale Sentinel and Newcom Media see the upside, but also stress the importance of ethnics, transparency, and human oversight. The big worry is fake "news" sites using AI to mimic journalism without any actual journalism, which is a credibility crisis waiting to happen - not missed by readers of the Gallup survey.
Read

Local news can't break through
After a tragedy in UK city Manchester, Google favoured US brands CNN and The New York Times over super respected and very local news source the Manchester Evening News, despite MEN having people on the ground at the scene. Experts say it is another sign that the supposed genius Google algorithm merely selects big outlets and side-lines the local voices, while a feature to pick preferred news sources exists only in the US and India for now. Local news fans, stay tuned.
Read

Phantom sources, real blunder
Deloitte Australia's recent report on welfare penalty automation has taken a nosedive into the land of AI, filled with hallucinated quotes and phantom research. The report, which cost the taxpayer $440K, leaned on generative AI which Deloitte seemingly forgot to mention upfront. Now they are offering a partial refund, although the government insists that the core recommendations still stand. In the end, this is yet another reminder that when AI joins the team, transparency is the real compliance check.
Read

Sora scrolls in. What now?
OpenAI has let Sora into the wild, an endless scroll of AI-generated mini videos starring different avatars as well as trademarked favourites. While some call it a flashy gimmick and others raise eyebrows at its appetite for copyrighted content, the apps rise is undisputable. Now, OpenAI is betting on monetisation and actually giving a say to rights holders (shocker, we know), but the real question is whether users are willing to trade reality for an infinite stream of AI fantasy.
Read

More money more chaos
Reddit, once home to memes, mayhem and questionable advice - it's where the glue on pizza tip originated - now gathers over $130 million a year from content deals with Google and OpenAI. What it also got is a handy traffic spike as well as the starring role in AI-generated search results. But it all comes with a headache, as AI has been swiping Reddit's content without permission and prompting the launch of its own AI search, suing crawlers, and the backing of proposed new licensing rules.
Read

Cookies crumble, first-party data rises
As publishers transition to a first-party data world and third-party cookies diminish in importance, big players such as CNN and Schibsted are already cashing in with ads programmes built on first-party insights. INMA reveals the "crawl, walk, run" methods being used by some early adopters to keep advertisers happy, users somewhat private, and publishers profitable.
Read