Download this opinion piece as a one-page PDF

Google Will Adapt. TripAdvisor Might Not.

A German court just did something whose implications almost everyone is reading backwards, and I want to make the honest, admiring case for why.

According to reporting on the decision, on May 28, 2026 the Regional Court of Munich I issued a preliminary injunction finding that Google can be held directly liable for false statements its AI Overviews produce, on the logic that an AI-written summary is Google’s own published content, not a neutral pointer to what other people wrote. The reported reasoning is that the model synthesizes, rewrites, and restructures information into new statements in its own voice, so it is speaking, not indexing, which meant Google could not shelter behind the hosting safe-harbor that has protected search engines for years. Google confirmed on June 12 that it is appealing.

The instant hot take was predictable: “this is a nightmare for Google.” I think that is exactly wrong. And once you see why, you also see who should actually be worried, and it is not the search giant everyone loves to cast as the villain.

What the Munich court did, and did not, do

Precision first, because this ruling is being wildly over-read in both directions.

What it did, per the reporting: in a case brought by two Munich publishing houses whose reputations were damaged by false AI-generated claims (that they ran subscription-trap scams and fraudulent schemes, claims the court found had no basis in the cited sources), the court treated the AI Overview as Google’s own statement and granted a preliminary injunction against repeating it, grounded in German business-personality-rights law. It reportedly rejected Google’s argument that users can just fact-check the answer themselves.

What it did not do, and this matters enormously: it did not issue a final merits judgment, it is under appeal as of June 12, it binds only those two plaintiffs, and, because Germany is a civil-law system, it sets no binding precedent even domestically. It is a persuasive, provisional, contested signal, not a settled law of the universe. Anyone selling you “AI Overviews are now illegal” is overselling a preliminary injunction. The honest read is narrower and more interesting: a serious court found the “an AI answer is the publisher’s own speech” theory persuasive enough to act on it, and that theory, if it spreads, reshapes the risk map for everyone who ships an AI answer.

So the real question is not “is Google doomed.” It is: if a standard like this spreads, who is equipped to handle it? And on that question there is no contest.

Whether you agree with the ruling is beside the point

You can think the Munich court got it exactly right or completely wrong. It does not change the thing I actually care about, which is capability. And Google is, by a distance, the best-positioned company on the planet to absorb a standard like this.

People picture Google as a lumbering target here. I see the opposite. AI Overviews are a choice, a feature Google elected to bolt onto a results page it has spent twenty-five years refining into the most valuable real estate on the internet. Every element of that page is a dial Google can turn. It can scale AI Overviews back to the queries where they are safest. It can lean harder into the citation-and-link format, an answer that visibly points to its sources rather than replacing them, which, not incidentally, is also the format that sits nearer the defensible end of the legal spectrum I will get to. It can add guardrails and confidence thresholds. It can, at any moment, fall back to the thing it is still better at than anyone alive: ranking and displaying directly-indexed links. It is already rolling out an AI Overviews opt-out in the UK under a regulator’s order. A company with that many exits does not get trapped in the room.

And here is the part independent hoteliers should genuinely appreciate, because it is easy to be reflexively cynical about Google and wrong. Google’s business model does not need to hurt you. It monetizes attention through advertising whether it shows an AI answer or ten blue links, which means, unlike an online travel agency, it does not have to intermediate your booking to win. It wins by being the front door. A Google that reshapes its results under legal pressure is still a Google that can, and commercially wants to, hand your future guest a path toward your own direct channel. Its incentives and yours are far more aligned than a decade of SEO paranoia gave it credit for. Adaptability plus aligned incentives is why, for Google, a liability standard is a formatting-and-engineering problem, expensive maybe, existential no.

I am not worried about Google. Google is going to be completely fine, and it is going to keep being the most important discovery surface an independent hotel has.

Would America even go there? The honest legal picture

Before I point at who should be worried, let me be scrupulously honest about the biggest caveat in this entire piece: nothing like the Munich rule exists in US law today, and it is genuinely unresolved whether it could.

Here is the real landscape, and I am going to resist the urge to overstate it. In the US, Section 230 of the Communications Decency Act broadly protects platforms from being treated as the publisher of content “provided by another.” The billion-dollar open question, which no US court has squarely decided, is whether an AI system’s own synthesized output counts as “another’s” content (protected) or the platform’s own developed content (not protected). Legal scholars, and the Congressional Research Service, frame generative AI as sitting on a spectrum: a retrieval engine that mostly surfaces third-party material sits nearer the protected end, while a “creative” engine that generates genuinely new statements sits nearer the unprotected end, because inventing claims that appear in no source starts to look like developing content yourself. The Supreme Court’s 2024 Moody v. NetChoice decision, separately, held that a platform’s editorial curation is its own protected speech, a double-edged sword: it shields platforms from being forced to carry speech, but its “curation is our own speech” logic can also be turned around to argue that heavy synthesis is the platform speaking. And the one real-world AI-defamation case to reach a decision so far, Walters v. OpenAI (a 2025 Georgia trial-court win for OpenAI), turned on ordinary defamation hurdles, disclaimers, the actual-malice standard, and lack of damages, not on Section 230 at all.

Put simply: the US does not have the European rule, a US plaintiff pushing this theory would face a steep, uphill fight, and I am not predicting the rule crosses the Atlantic. But the doctrinal seeds for a court to get there already exist, which is why this is worth thinking about now rather than after. Everything that follows is a hypothetical: if a standard like this took hold in the US. With that firmly stated, notice what the spectrum implies. A retrieval-and-citation answer that points to sources is more defensible than a fully synthesized one. That is a structural advantage for the player whose answers can most easily cite and link, and a structural exposure for the player whose entire product is synthesis.

The zombies have nowhere to pivot

Now turn to the businesses that do not have a healthy core to retreat to: the aggregators and review indexes that were already being hollowed out by the shift to AI answers, and that responded not by adapting from a position of strength but by betting the farm on becoming AI-answer machines themselves. The industry zombies. And in hotels, the flagship example, I am sorry to say, is TripAdvisor.

I have written before about how TripAdvisor’s core relevance is quietly draining, in Is TripAdvisor Dying? An SEO Post-Mortem in Progress and in the look at how to break its ‘best-of’ moat on Google. Its strategic answer has been AI-generated review summaries, launched in 2023, a synthesized paragraph at the top of a hotel’s page, built with retrieval-augmented generation over its own review corpus, telling you what “travelers” think. Structurally, it is the same move Google made, with one enormous difference: for Google the AI summary is one feature among many. For a business whose whole remaining value proposition is trustworthy aggregated review authority, the AI summary is closer to the whole bet.

And that is where it gets uncomfortable, because of two things stacking on top of each other.

First, the legal geometry. TripAdvisor’s AI summaries do not sit on Google’s page. They sit on TripAdvisor’s own hotel pages, generated in TripAdvisor’s own synthesized voice, about specific named businesses. In the US spectrum I described, a product whose core function is synthesizing reviews into new evaluative prose sits nearer the “creative synthesis” end, the exact zone where Section 230’s protection is least tested and where the “you developed this content, so it is your speech” argument is strongest. I want to be careful here: this is my analysis, a contested inference, not a settled legal conclusion, and there is a real counterargument that a faithful aggregation of user reviews is just a neutral compilation. But if AI-answer liability ever does come to America, a review platform’s synthesized summaries look, to me, like a more exposed target than a search engine’s cited, link-out answers.

Second, and this is what turns exposure into an emergency, is what a national consumer watchdog just reported about those summaries. In July 2026, the UK consumer group Which? published an investigation reporting that some of TripAdvisor’s AI review summaries downplayed serious safety concerns that were present in the individual reviews. According to Which?, a summary for one Cape Verde resort (the Riu Palace Santa Maria) described it as “spotless” and its rooms as “spacious,” while, per Which?, the underlying reviews contained scores of food-poisoning complaints and the property was tied to a group legal action by holidaymakers who say they fell ill. According to Which?, a summary for a hotel in Turkey (the Kaia Coracesium) described service as “friendly” with only minor “lapses,” while, per Which?, reviews alleged repeated sexual harassment by staff. I want to be precise: these are Which?’s findings and the allegations in reviews and legal proceedings it cites, not facts established in court, and I am not asserting any of them as proven. Which? also reported running the same hotels through Google’s AI summaries and finding Google’s more cautious, in the Cape Verde case, per Which?, Google’s answer flagged the “potential for illness.”

TripAdvisor strongly disputes all of this, and fairness requires its full response. In the statement it gave Which?, TripAdvisor said: “We fundamentally disagree with the premise of this investigation. Our AI Summaries have been designed to uphold the integrity and transparency that has made Tripadvisor trusted by millions of travelers for over 25 years. They provide snapshots based on high volumes of user generated content and explicitly are not intended to replace individual reviews.” It added that it has “comprehensive safeguards in place,” that its systems “automatically suppress AI Summaries for listings that feature warnings from travellers about serious safety incidents such as death, drugging or sexual assault,” that it is “actively looking into” the examples raised, and that the suggestion its tools “pose danger to travelers is an unfounded claim that seems designed to generate controversy rather than inform readers.” You should weigh that response seriously; if its suppression safeguards work as described, that is a meaningful rebuttal to the specific examples.

But sit with the shape of it anyway, because it is the whole essay. The company whose entire remaining reason to exist is trustworthy aggregated review authority is leaning its future on an AI product that a reputable national watchdog has publicly criticized on safety grounds, in a legal climate that is, at minimum, starting to ask whether the publisher of an AI answer owns its words. That is the definition of a business-model apocalypse looming, and, unlike Google, TripAdvisor cannot simply walk away from AI summaries, because they are not a feature it added on top of a healthy business. They are the reinvention it is counting on. You cannot retreat to a strategy you have already committed to as your future.

That, to me, is the difference between an adaptable giant and a zombie. Google has ten doors out of the room and an incentive to keep serving hotels on the way out. A platform that bet its relevance on synthesized review answers is nearer the fire, with fewer exits.

What this actually means for an independent hotel

I did not write this to dance on anyone’s grave, though I will admit it is not the least satisfying essay I have written. I wrote it because the strategic lesson for an operator is clear.

Keep betting on the adaptable, aligned platform. Whatever format Google settles into, AI answers dialed up, dialed down, wrapped in citations, or quietly rested for a season, it remains incentivized to be a good front door and to route travelers onward, including to your direct site. So keep the surfaces it rewards in excellent shape: a complete Google Business Profile, clean structured data, a steady flow of genuine reviews, and real answer-engine presence. You are investing in the player with the most options and the incentives closest to yours.

And stop over-feeding the zombies. Every review and every ounce of reputation you pour into a platform whose AI product is simultaneously its bet for survival and, potentially, its biggest liability is a fragile investment. Move your review gravity to the surfaces you and Google both trust, which I laid out in the drain-the-moat piece, and own accurate, complete first-party signals about your property everywhere a machine can read them.

Because here is the quietly optimistic punchline for the honest operator. Accountability for AI answers, whether it arrives through courts, regulators, or watchdog pressure like Which?’s, is good for you. A world where an AI summary cannot casually gloss over a hotel’s real problems is a world where your genuinely good hotel’s real signals, real reviews, real cleanliness, real service, count for more, not less. The reckoning that unsettles the zombies is the one that finally rewards the operators who actually earned their reputation. I would bet on that world. And I would bet on Google helping build it.

The honest caveats, in full

This is opinion, and I would rather you trust the reasoning than the rhetoric, so hold every limit in view.

The Munich decision is a preliminary injunction, it is German rather than US law, it binds only the two plaintiff publishers, it sets no binding precedent even in Germany, and Google is actively appealing it, so it may narrow, be overturned, or never migrate anywhere. Whether anything like it reaches American courts is genuine speculation on my part; US law today does not have this rule, and the Section 230 question for AI output is unresolved, not decided. My argument that Google is better positioned than TripAdvisor, and that synthesized review summaries are more legally exposed than cited search answers, is an argument from disclosed facts and legal analysis, not a prediction or a legal conclusion, and reasonable lawyers would push back on parts of it. Every factual claim about TripAdvisor’s AI summaries is attributed to the Which? investigation and reporting on it; those are Which?’s findings and the underlying allegations, not adjudicated facts, and TripAdvisor’s full rebuttal is quoted above. I am not stating that TripAdvisor has done anything unlawful or that anyone has been found liable. And, to say it once more, none of this is legal advice; for that, talk to a real lawyer, not a hotel SEO blog.

What survives all of that, I think, is the shape of it: the platform with the most optionality and the most aligned incentives will adapt and keep serving hotels, and the platforms that bet their survival on an AI product they may not be able to stand behind are the ones with more to lose. That is my opinion. Go verify it.

If you want help pointing your discovery strategy at the durable, aligned surfaces instead of the fragile ones, the Google Business Profile and answer-engine work, the review-gravity shift, the first-party signal cleanup, that is the work we do at HotelSEO Lab. You can see how we break it down on our services page. No doom-selling, no guarantees, just an honest read on where the ground is solid and where it is smoking.