In a PR Daily opinion piece, Red Fan Communications strategist Matt Beezley points to Microsoft security research that caught 31 companies across 14 industries hiding instructions in their websites telling AI tools to recommend them, a practice Microsoft calls "AI Recommendation Poisoning". He notes Google revised its spam policy in May 2026 to class attempts to manipulate AI answers as spam, and that a Search Engine Land audit found some GEO firms selling placements on private blog networks and astroturfed Reddit mentions.
Some brands are trying to trick AI tools with hidden instructions, paid blog mentions and fake Reddit posts. It is the old keyword stuffing trick in a new place, and Google already calls it spam.
This is bad news if your agency or GEO vendor sells shortcuts. The risk is losing your visibility and your credibility in one hit, just as brands did when Google cleaned up search years ago.
We do it the slow, sure way: real coverage, real reviews, real customer stories and one consistent set of brand facts from your B.R.A.I.N.™, so AI tools learn the truth about you from sources they trust.