WebMCP Can Be Used To Hijack AI Agents, Chrome Warns via @sejournal, @martinibus

<p>The agentic web shipped for real in April. Here's what Cloudflare, OpenAI, and Sundar Pichai all said that web professionals need to hear.</p> <p>The post <a

<p>Google warns that websites can expose AI agents to hidden traps for agentic AI navigating the open web.</p> <p>The post <a

<p>AI agents are getting good at using the web. But the way they interact with it today is fragile: CSS selectors, XPath queries, visual parsing, and DOM scraping that breaks

<p>OpenAI added WebMCP support to ChatGPT’s desktop browser, letting supported webpages expose structured actions to ChatGPT Work and Codex.</p> <p>The post <a

<p>Agentic browsers and commerce protocols are redefining how users interact with websites, and who actually completes transactions.</p> <p>The post <a

New Microsoft&nbsp;research&nbsp;shows how attackers can hijack AI agents that act on a user's behalf, using nothing more than a poisoned tool description to make the agent

🤖 This week Google announced a new user agent just for agents - @mariehaynes.bsky.social covered this release in a must read piece: Why Google’s New “Google-Agent” is the Biggest

<p>Apple's WebKit announced a new MCP server for Safari that enables AI-based website debugging for SEO and Core Web Vitals.</p> <p>The post <a