Source unites Oxylabs and Decodo: AI agents are reading your store

The biggest names in web scraping just formed one group. Oxylabs, Decodo, Webshare, Honeygain, and ScrapingBee now operate under a single name, Source. The group says AI agents and AI search drove a four-fold rise in demand across its companies in a year. If you sell online, your product pages, prices, and reviews are some of that public web data. It is time to decide which AI bots get your data, and on what terms.

What actually happened

Source announced its launch from Vilnius on October 13 in a GlobeNewswire release. The facts, as the company states them:

• Five brands, one group. The brands keep operating independently. Oxylabs serves enterprises and AI labs, Decodo the mid-market, ScrapingBee developers through its API, Webshare prosumers, and Honeygain consumers who share their spare bandwidth.
• Demand: a four-fold increase in demand in the last year, which the group links to next-generation search, AI agents, and frontier AI products.
• Scale: 800 people across Europe, the U.S., and Asia, more than 350 of them engineers, and a $3.6 billion valuation from a Warburg Pincus-led round this year.
• The message: CEO Vytautas Savickas said the new name gives the group "a unified voice in defending responsible access to public web data."

That last line is the one to note. A scraping industry with one voice will push back on anyone who tries to close the public web.

Why it matters for your business

Some of this traffic is good for you. A shopping agent that reads your catalog can send you a buyer. A competitor's price tracker that hits your product pages every five minutes cannot. The two can look the same in your logs, and the bandwidth-sharing model behind Honeygain means a lot of scraper traffic can come from ordinary home IP addresses. Blocking by IP will not sort it.

So make the decision on purpose:

• Look at your bot traffic. Your CDN or host can show you which user agents hit you and how often.
• Publish a policy. Set robots.txt rules for the AI crawlers you know, and allow the shopping and search agents that send buyers.
• Rate-limit the expensive endpoints. Search, inventory, and pricing APIs should not answer unlimited anonymous requests.
• Give agents a clean feed. Structured product data lets the agents you want get accurate prices without scraping every page.

Sources: Source via GlobeNewswire, GlobeNewswire release.