“As AI agents become active internet users rather than passive scrapers, websites and tech companies are developing new standards and access tiers specifically for them. This shift, driven by protocols like Anthropic's Model Context Protocol and emerging 'agent-friendly' web layers, risks creating a two-speed internet. The implications for web openness, data access, and human browsing experience are still poorly understood.”
Key Takeaways
- Websites traditionally split users into bots and humans, but AI agents now occupy a disruptive middle ground requiring new classifications.
- Anthropic's Model Context Protocol (MCP) is one emerging standard designed to let AI agents interact with web services more efficiently.
- Optimising the web for AI agents may degrade the experience for human users, raising accessibility and equity concerns.
A new internet tier for AI agents could quietly erode the open web for humans.
trending_upWhy It Matters
If a parallel, agent-optimised web layer emerges, large AI companies with resources to negotiate API access could gain outsized advantages over smaller players and independent developers. Human users may find websites increasingly designed around machine interaction patterns rather than their own needs. Regulators and standards bodies have not yet caught up with this shift, leaving the rules of the road largely written by the companies building the agents. Watching how bodies like the W3C and governments respond to agent-specific web protocols will be critical in the next 12-24 months.
FAQ
What makes AI agents different from the web bots websites already handle?
Traditional bots perform narrow, repetitive tasks like indexing pages, while AI agents make dynamic decisions, fill forms, click links, and transact on behalf of users. This broader, more human-like behaviour does not fit neatly into existing bot-management frameworks like robots.txt.
What is the Model Context Protocol and who created it?
Model Context Protocol (MCP) is an open standard created by Anthropic that allows AI agents to interact with external tools and web services in a structured way. It is designed to give agents reliable, efficient access to data and actions across different platforms.
How could optimising for AI agents harm everyday internet users?
Resources spent building agent-friendly interfaces may come at the cost of human-centred design, accessibility features, and page performance. There is also a risk that premium agent-access tiers create paywalled web content that human users, and smaller AI developers, cannot reach.
