AI agents are expected to research the internet independently, gather information and carry out complex tasks. But an increasing number of websites are restricting their access. As The Verge reports, AI agents from OpenAI, Meta and other providers are encountering growing technical and legal barriers.
Companies such as Amazon, Yelp and Delta Air Lines are limiting access for automated AI systems. While Yelp links access to paid data licensing agreements, financial services providers such as JPMorgan Chase and Charles Schwab block agents from accessing customer accounts for security reasons. Existing anti-bot systems can also block AI agents, even where they are not specifically intended to do so.
This development is particularly relevant to professional research: AI agents can only conduct reliable independent research if they are able to access the information sources they need. Increasing access restrictions could therefore reduce the completeness of automated research and limit the scope for continuous information gathering.
At the same time, new models for controlled access are emerging. Yelp, for example, allows certain AI agents to access its data through licensing agreements, while Meta is entering into partnerships with various companies. This development highlights a fundamental tension between the automation of digital workflows and website operators’ interest in controlling access to their content and services.
For companies, this means that the use of AI agents should take into account the availability of relevant data sources, potential access restrictions and the reliability of automated research processes.
By Maria Kleiner
Source: The Verge, 17 October 2026, by Emma Roth