The Agentic Web Is Here—and It Wasn't Invited

AI agents are rewriting the web's basic logic. The question isn't whether it changes—it's who controls the outcome.

By The energy-climate Desk

A quarter-century later, that vision has arrived, just not as designed. New tools such as Meta's blockbuster Muse agent and OpenAI's Dots are logging record downloads and being given access to people's bank accounts and email inboxes [1]. The agentic web is no longer a thought experiment.

Unlike Berners-Lee's semantic agents, today's LLM-based tools require no special tags to navigate a website—they read, click, and transact across the open web the same way a human would, only faster [1]. Time magazine is already publishing versions of its articles that are easier for agents to parse and has started selling ads targeted at them. Companies from Etsy to Expedia have partnered with the major AI firms to build tools that allow agents to interface seamlessly with their services [1].

GPT-6 Astra achieved a 73% success rate on OSWorld 2.0, a benchmark for complex, multistep computer-use tasks [1]. The human web's defenses are not keeping pace.

Two futures sit at the fork. One opens the web wider—cheaper, more direct, free of algorithmic middlemen. The other concentrates it further, routing daily life through a handful of agent platforms whose incentives users cannot audit or override.

The attention economy's foundations are already cracking. Who pours the new ones matters enormously.