April 09, 2026 • SEL • Important*
2026-04-09 • SEL • Important
What Happened
A new report confirms what publishers have been feeling in their analytics: AI bot traffic surged 300% in 2025, and the bill is being sent to content creators. The data shows two distinct categories of bots — training crawlers that scrape your content to build AI models, and fetcher bots that pull your content to serve instant answers — both growing at rates that outpace legitimate human traffic.
For publishers, this creates a double squeeze. Training bots consume your content without permission or compensation, building commercial AI products on the back of your reporting. Fetcher bots surface your content in AI Overviews and chat responses, often without sending a click back to your site. The result: your content works harder for everyone except you.
Key Takeaways
- Two bot categories, two different threats. Training bots (63% of AI bot traffic in publishing) harvest content for model development. Fetcher bots (24%) pull content for real-time AI responses. Both reduce your traffic. Only one might send you a citation.
- Pay-per-crawl is gaining traction. Some publishers are moving toward models where bots must pay to access content. This is a structural shift from the open web model that's existed for decades.
- Selective blocking works better than blanket bans. Blocking all bots risks losing AI Overview citations and the traffic they still generate. The smarter play: allow fetchers that attribute, block scrapers that don't.
- Monitoring matters more than ever. Most businesses have no idea how much bot traffic they're receiving. If you're not measuring it, you can't manage it.
The ONmetrics Take
This isn't a future problem — it's a current one. Here's what London, Ontario businesses should do about it:
1. Check your server logs. Use your hosting analytics or Cloudflare dashboard to see what percentage of your traffic comes from known AI bot user agents. You might be surprised at the volume.
2. Don't block everything. Google's fetcher bots can still send traffic through AI Overviews. The bots to block are the ones that scrape without attribution — check your robots.txt and consider selective allow/deny rules.
3. First-party data is your moat. AI bots can scrape your blog posts. They can't scrape your customer insights, proprietary data, or local market knowledge. Content that draws on what only you know can't be replicated by a crawler.
The bottom line: AI isn't replacing SEO — it's raising the bar on what ranks. The content that survives the bot surge is the content that offers something a crawler can't copy.
Source: SEL