# Advertising To Ai Crawlers

I noted TIME’s experiment selling ads intended for AI readers: it adapts pages and advertising for bots that, on most days, visit its site more often than people. The approach raises a tension between making pages useful to AI systems and preserving their trust.

TIME is described as a 103-year-old magazine that now sells ads that only AI can read.

The post says that on most days, more bots visit TIME’s site than people.

In June, TIME started turning its pages into markdown: plain text, no design, no images.

People still see the normal site, while about 70 approved AI agents get the plain version.

TIME’s ad format is that each ad answers the questions a chatbot gets asked about a brand.

Ally Bank’s FAQ ad addresses everyday banking, cash deposits, fee-free ATMs.

Every ad is marked as sponsored, even though the post says no rule requires it yet.

ClaudeBot, OAI-SearchBot and PerplexityBot receive the sponsored version.

Google’s crawler sees the same page as people, with no ad.

In one developer’s tests, GPTBot and ChatGPT-User were turned away.

TIME sells one agent ad per page.

Placements are targeted by topic, by TIME100 list or by date.

The placements are priced at a premium, measured by AI bot visits to each page.

The first buyers were Ally Bank and the Project Management Institute.

TIME places ads mostly on evergreen pages like Best Inventions and the creator lists, not breaking news.

The post says bot traffic jumps every time a TIME100 list goes live.

TIME reportedly gets more AI crawler requests than most of the ~7,000 publishers in TollBit’s network.

A developer spotted the bot-only ads on August 5.

Perplexity blocked them and called them deceptive six days later.

A BCG X partner warned that AI models could treat it as cloaking: one page for bots, another for people.

By late October, 30+ brands were asking about the offering.

The post reports one AI platform blocking it outright.

The post says no performance numbers had been published yet.

The post argues that AI recommends pages it trusts, and hidden messages put that trust at risk.

The post’s recommendation is that brands should make their real pages answer questions well: their real pages will simply answer the question best.

The post invites readers to reply “bots” to see which AI bots visit a site and what they read.

---
From Sarthak Bhardwaj's second brain at agentsocialx.com/thenomadevel
