Something rare happened in search this year: Google and Microsoft agreed. Within months of each other, both companies published official guidance on how websites appear inside AI-generated answers, and both delivered exactly the same uncomfortable message to the shortcut industry. The tricks do not work. Google’s guide dismisses them one by one. Bing went further and reclassified some of them as spam. For businesses that spent two years chasing AI visibility hacks, the ground shifted sharply beneath them.
What the engines actually said
Google’s Search Central team published its official guide to optimising for generative AI features in May, and it reads like a systematic dismantling of popular advice. There is no special schema.org markup that earns AI citations. llms.txt files, promoted widely as the new robots.txt for AI, are ignored by Google Search entirely; they neither help nor harm. Breaking content into bite-sized chunks for machines is not required. Manufactured brand mentions across forums and blogs fail because the same spam systems that police regular search also govern AI features built on top of it. Google’s position on AEO and GEO as disciplines is blunt: optimising for generative AI search is still SEO.
Bing’s rewrite of its Webmaster Guidelines took a different route to the same destination. Microsoft named Generative Engine Optimisation as a legitimate practice, gave publishers real controls over how Copilot cites their content, and shipped an AI Performance report in Bing Webmaster Tools that shows citations across Copilot and Bing’s AI summaries. Then it expanded its abuse definitions to address the new terrain. The section once titled keyword stuffing now covers artificially engineered language, meaning content written to bait AI citations rather than inform readers. What used to be a grey-hat tactic is now a named spam category within one of the two major search engines. The message across both documents is consistent: the engines will measure your visibility for you, and they will punish attempts to fake it.
Why the hacks were always going to fail
The convergence is not a coincidence. Both engines build AI answers the same basic way: they retrieve passages from indexed pages, weigh the evidence, and compose a response with citations pointing back to sources. Bing’s engineers describe the shift plainly. The unit of value in search is moving from whole documents to discrete, supportable facts with clear provenance. A system built to verify claims before repeating them has little use for pages that contain clever formatting but nothing worth verifying.
That is why the hacks fail. An llms.txt file adds no facts. Citation-bait phrasing adds no provenance. The machinery these tactics try to impress looks straight past them and focuses on the substance, rewarding whoever supplies it.
Not everyone accepts Google’s framing, and the disagreement is worth understanding. Several respected practitioners have argued the guidance is self-serving, pointing out that retrieval systems genuinely do work on passages, so how content survives being split into fragments matters more than Google admits. Bing’s own documentation concedes the point, noting that content transformations must preserve the meaning of claims used in answers. The dispute, though, is about mechanics, not morals. Even Google’s sharpest critics are arguing for clearer, more self-contained writing. Nobody credible is defending the tricks.
What a hack-free content strategy looks like
Strip away the shortcuts and the work that remains is demanding but manageable. This is the discipline behind Generative Engine Optimisation done properly, and the reason serious practitioners treat answer engine optimisation as an extension of content quality rather than a separate bag of tricks.
Four habits do most of the work. State your central claim in the opening paragraph of every page, so a retrieved passage carries its own meaning. Support claims with facts you can prove: dates, figures, named sources, first-hand experience. Attach real authorship, because verifiable expertise carries more weight than anonymous authority. Keep your entities consistent, so your business name, services and locations match across your site, your profiles and the wider web.
None of this is new. Any capable SEO agency Melbourne businesses rely on has already been building this way for years, which is exactly Google’s point. The engines did not invent a new game. They raised the stakes on the old one.
The advantage goes to the rigorous
Both engines have now published the answer key. That removes the mystery and moves the contest to execution: who can produce provable, well-structured, genuinely useful content with the greatest consistency. That is a contest a disciplined business can win.
If you want an honest read on where your content stands, Myoho Marketing offers a free GEO consultation that reviews your site the way an answer engine does: claims, provenance, structure and authorship. As a digital marketing agency Melbourne businesses have trusted for more than a decade, we build strategies on the fundamentals the engines just confirmed. Book your free consultation today.
Thank you for entering the lucky draw! We wish you all the best. Winners will be announced by email.