Akamai analyzed AI bot activity by examining application-layer traffic from its bot management tools.
Commerce drew the most AI bot traffic at 48%. Media, which includes publishing, video, social media, and broadcasting, came second at 13%.
Publishing companies accounted for 40% of all AI bot activity in media, ahead of broadcast and OTT at 29%.
OpenAI generated the most AI bot traffic hitting media companies, with 40% of its media requests going to publishing companies. That’s partly because OpenAI runs multiple bots. GPTBot handles training, OAI-SearchBot powers AI search, and ChatGPT-User retrieves content in real time.
Meta and ByteDance were the second- and third-largest operators. Anthropic and Perplexity rounded out the top five at lower volumes.
Why Akamai Says Fetcher Bots Are The Bigger Concern
The report groups AI bots into four types based on behavior.
Training crawlers and fetchers account for most of the AI bot activity Akamai saw in media, which includes publishing. Training crawlers collect content to build language models. They made up 63% of AI bot activity targeting media in H2 2025.
Fetcher bots grab specific pages in real time when someone asks an AI chatbot a question. They made up 24%, and publishing accounted for 43% of that fetcher activity.
Akamai argues that fetcher bots are the more immediate revenue concern, even though training crawlers generate more total traffic. When a fetcher bot pulls an article to answer a chatbot query, the user gets the information without visiting the publisher’s site.
How Publishers Are Responding
It’s worth noting that Akamai sells bot management tools, and the report’s recommendations point toward its own products and partners.
The most common responses among Akamai’s customers are deny (blocking requests outright), tarpit (holding connections open to waste bot resources), and delay (adding a pause before responding). One unnamed publisher chose tarpitting over blocking, controlled 97% of AI bot requests, and kept the door open to potential licensing deals.
The report argues against blanket blocking, saying some AI companies are willing to pay for content access and that blocking all bots removes that option.
Looking Ahead
The report’s top takeaway is the distinction between training crawlers and fetcher bots. Blocking a training crawler can influence how your content helps build future AI models. Blocking a fetcher bot affects whether your content appears in AI responses right now.
Featured Image: la pico de gallo/Shutterstock
