# MIT Technology Review Brasil: adapting the web to AI agents could harm humans

- Published: 2026-10-25T09:00:03.000Z
- Updated: 2026-10-25T10:01:48.498Z
- Section: Market
- Topics: Agents, Society
- Page: https://aidailyjournal.com/en/n/2026/10/25/mit-technology-review-brasil-adapting-the-web-to-ai-agents-could-harm-humans/
- Sources: [MIT Technology Review Brasil](https://mittechreview.com.br/agentes-de-ia-podem-tornar-a-internet-pior-para-humanos)

## Summary

MIT Technology Review Brasil says personal AI agents, which can book trips, negotiate prices and make purchases on users’ behalf, are putting pressure on a web designed mainly for humans. The article cites the launch of OpenAI’s Dots and Meta’s Muse; according to the publication, Muse had 5 million downloads in its first month and AI-agent traffic tripled from August to September.

The article reports that sites including Pizza Hut, Yelp and Amazon have been blocking agents, while Amazon is engaged in a legal dispute with Perplexity. Ethan Mollick of the Wharton School predicts that using these agents to request refunds or negotiate prices could overwhelm customer service. The publication also notes that Shopify already offers agents dedicated checkout channels, and questions how these changes will affect other internet users.

## Why it matters (editorial interpretation)

The adoption of agents that handle purchases and customer service could change how sites control access and increase pressure on support operations, according to the publication and Ethan Mollick. The clash among blocking, dedicated channels and Amazon’s legal dispute with Perplexity points to new limits for intermediation on the web.

_Summary written by AI Daily Journal in its own words, based on the sources above. For details, read the original articles._