Source [De]Coded —Publishers are rethinking Google, and it changes how AI describes you
Author
Greg Mondshein & Becky Honeyman
07.30.2026
Hey team!
While half the industry seems to have mentally checked out for July, we are operating at full speed. I just published a piece in PRWeek on why the summer slowdown is a complete myth and how smart CMOs use this window to win Q4. We also just launched our new Brand Equity ROI Multiplier whitepaper alongside IMS to show marketers how brand value actually turns into revenue, and if you are on the West Coast, come hang out at our upcoming SF Summer Soiree on August 25. It is amazing what happens when you put the right team in place and give them the space to crush it.
Speaking of staying in control, keeping your hand on the wheel is getting tricky across tech, media, and marketing. This week, we examine why the UK shuttering its dedicated tech department risks stalling sovereign AI momentum, how Coca-Cola is using AI for global brand governance instead of just generating more noise, and why major publishers are finally asking if feeding Google search is still worth the trade. Plus, we look at how a single bad review turned into a national crisis when a manager broke the ultimate rule of brand protection.
Let’s get into it.
Greg and Becky
Closing DSIT could stall the UK’s ambition for sovereign AI
Scrapping a dedicated technology department risks trading policy momentum for bureaucratic reshuffling at a critical moment for UK tech.
TL;DR: The UK government’s decision to dissolve the Department for Science, Innovation and Technology (DSIT) and merge its functions into a broader business ministry risks creating a 6-to-12-month policy vacuum. SourceCode Communications argues that while folding tech into a growth mandate makes sense on paper, the resulting administrative friction threatens the UK’s pitch as a stable, attractive destination for sovereign AI investment.
Takeaway: Government re-organizations rarely feel urgent to tech companies until the policy certainty they relied on quietly vanishes. DSIT functioned as a single point of entry for digital regulation, compute access, and AI security. Absorbing it into a wider apparatus creates a classic operational risk: official attention turns inward to org charts and briefs just as international competition for capital accelerates. For tech brands, public sector partnerships, and institutional investors, the primary risk isn’t a policy change—it’s the friction and delay of nobody knowing who owns the table anymore.
Consider:
- For marketers, if policy shifts and department mergers disrupt the timeline for national tech initiatives, how flexible is your messaging when anchored to public sector momentum or government partnerships?
- For communications leaders, as digital governance moves across departments, do you have direct relationships with the policy stakeholders who actually hold the pen, or are you still relying on legacy access points that no longer exist?
Coca-Cola built an AI system to keep itself on brand across 200 markets
Most marketers are pointing to AI at making more work. Coca-Cola pointed it at making the work consistent.
TL;DR: Coca-Cola has deployed an AI system called Project Fizzion, built with Adobe, to govern brand consistency across roughly 200 markets. Rather than generating more assets, it checks and guides work against the company’s brand standards. Forbes frames it as a governance model other marketing organisations can borrow.
Takeaway: Almost every AI marketing story of the past two years has been about volume. This one is about control, which is the more interesting problem. Brand consistency at global scale has always been an expensive coordination exercise held together by decks, guidelines nobody reads, and a great deal of email. If the durable use of AI inside large marketing organisations turns out to be enforcement rather than production, the teams that benefit will be the ones who have already written down what their brand actually is in a form a machine can check. Most have not.
Consider:
- For marketers, if an AI system had to audit your last hundred published assets against your brand standards, are those standards specific enough to actually run?
- For communications leaders, who owns brand governance in your organisation today, and do they have any say in which AI tools the teams producing the work are adopting?
Publishers are starting to ask whether Google is still worth it
Reddit, Politico and others are weighing whether to keep feeding a search engine that increasingly answers the question instead of passing it along.
https://www.wsj.com/business/media/google-search-publishers-ai-content-0fb06e41
TL;DR: Major publishers including Reddit and Politico are reconsidering their relationship with Google as answers produced by AI absorb traffic that used to arrive as clicks. Some are weighing whether they can work with the company at all. The Wall Street Journal reports that the arrangement which made Google indispensable to publishers no longer looks obviously worth it.
Takeaway: This is the clearest signal yet that the traffic bargain holding up the open web is being renegotiated in public. For anyone whose marketing leans on earned media, the downstream effect matters more than the headline: if publishers restrict what AI systems are allowed to read, the pool of material those systems use to describe your brand becomes smaller and more selective. Being visible inside an AI answer will increasingly depend on which sources are still willing to be read at all. That is a media relations question long before it is a technical one.
Consider:
- For marketers, do you know which publications the AI answers in your category are actually citing, or are you still grading coverage on impressions?
- For communications leaders, if the outlets you place in begin blocking AI crawlers, does your earned coverage still reach the places where buyers now ask their questions?
A bad review became a national story because a manager decided to answer it
The reputational damage was never the review. It was the retaliation.
https://www.autonews.com/retail/an-nissan-dealership-tiktok-review-apology-0723
TL;DR: The owners of a Florida upholstery shop documented a poor experience with a Nissan dealership and said they were then buried under a wave of negative reviews. The dealership apologised after its general manager was found to have retaliated against the customer. Automotive News covered the escalation.
Takeaway: Most organisations have a crisis communications plan and almost none have a plan for the moment a frustrated middle manager opens a review platform. The original complaint was survivable and very likely invisible. The response is what created the story, and it came from somebody who was never trained as a spokesperson but functionally became one. In most companies reputational risk sits with the people closest to the customer and furthest from the communications team, which is exactly backwards from where the policies live.
Consider:
- For marketers, who in your organisation currently has the ability to reply publicly on behalf of the brand, and has anybody reviewed that list this year?
- For communications leaders, does your escalation path cover a bad review, or does it only switch on once a journalist calls?