Skip to content

164news.com

  • Home
  • About Us
  • Contact Us
  • Privacy Policy
  • Terms of Service
  • Cookie Policy

Four teams just broke AI agents four ways in ten days. The flaw is the same one.

Posted on July 20, 2026 By 164news66 No Comments on Four teams just broke AI agents four ways in ten days. The flaw is the same one.

AI Agent Security: Four Attacks in Ten Days

In barely ten days, four separate research teams broke AI agents in four different ways. The pattern underneath them is the same, and it should worry anyone plugging a chatbot into their inbox.

July 20, 2026 – 2:09 pm

Image by: Canva

We spent two years asking whether AI would lie to us. The more useful question this summer is what happens once we hand it the keys. Give a model your Gmail, your calendar, a memory, and the power to act, and its mistakes stop being embarrassing—they start being exploitable.

Over roughly ten days in July, four pieces of research landed that make the same point from four angles. None is a lone bug; together they sketch the shape of AI agent security in 2026, and the picture is not reassuring.

The Browser Assistant That Clicks for You

Start with the tool sitting closest to your data. Manifold Security published research showing that any browser extension you install can quietly hijack Anthropic’s Claude for Chrome and make it read your Gmail, Google Docs, and Calendar.

The trick is small: The extension listens for a user click before it runs one of nine built-in tasks. It never checks whether the click is real. A rival extension can forge one in six lines of code, and Claude treats the fake as genuine.

Manifold rates it high severity by default. It jumps to critical once a user has switched on “Act without asking” mode, where the read happens silently. Manifold says it told Anthropic in May. Eight releases later, it reports, the flaw still works in the current version. The company had already covered earlier agentic browsing risks, and this is the same wound reopened.

A Single Email That Rewrites Memory

The second finding moves from clicks to memory. Personal agents now keep notes on you between chats: your contacts, your habits, your to-do list. Researchers asked what happens if an attacker can write to that store.

In a paper posted to arXiv, a team showed that one carefully written email can plant a false memory in an AI agent. They connected a victim inbox to an agent through Google sign-in, then sent payloads from an account they controlled.

The mail passed normal spam filters and reached the inbox. The agent read it, and in more than half of cases, it saved the attacker’s instructions into long-term memory with no warning to the user. That is the dangerous part: A normal prompt injection lasts one conversation. This one persists. The poisoned memory keeps steering the agent across future sessions until someone finds it. As Digital Trends put it, the agent remembers the lie and never mentions where it came from.

A Backdoor for Under $100

The third crack sits deeper still, inside the model itself. Katie Paxton-Fear, a cybersecurity lecturer and advocate at Semgrep, wanted to know whether you can trust an open-weight model you download for free. So she poisoned one.

It took about an hour and less than £75, she and two Semgrep colleagues wrote in a detailed post. Just ten tainted training examples were enough to make the model write code with a hidden security hole, even for prompts it had never seen before.

Clock

Post navigation

Previous Post: Netflix used AI in 300 shows. Its biggest director calls AI a Trojan horse.
Next Post: EVs outsold petrol and diesel in Germany for the first time. The birthplace of the car just changed its fuel.

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

Editor's Picks

  • Bronx DWI Defense Attorney
  • Brooklyn Small Business Legal Advice
  • New York Personal Injury Attorney
  • NYC Construction Law Specialist
  • NY Criminal Defense Lawyer
  • Manhattan Family Law Specialist
  • Queens Immigration Lawyer
  • Staten Island Civil Rights Advocate
  • Manhattan Personal Injury Prevention
  • Bronx Intellectual Property Attorney

Recent Posts

  • Protecting Assets: High Asset Divorce Attorney Worcester Guide
  • Family of Five Found Dead After Storms Flooded Utah Canyon
  • Protecting High-Value Assets: Worcester Divorce Attorney Strategies
  • Trump Relents From Spain-Bashing to Present Its Team the World Cup Trophy
  • Protecting Assets: High Asset Divorce Attorney Worcesters Guide

Recent Comments

  1. houseofluckcasino on Mick Jagger says AI is fine for musicians, as long as it does not sound like him
  2. happybingo on Mick Jagger says AI is fine for musicians, as long as it does not sound like him
  3. codwin on Mick Jagger says AI is fine for musicians, as long as it does not sound like him
  4. c444gameapk on Google DeepMind launched an AI biosecurity programme to fight biological threats
  5. bunny casino on Google DeepMind launched an AI biosecurity programme to fight biological threats

Archives

  • July 2026
  • June 2026
  • May 2026
  • April 2026
  • March 2026

Editor's Picks

  • Bronx DWI Defense Attorney
  • Brooklyn Small Business Legal Advice
  • New York Personal Injury Attorney
  • NYC Construction Law Specialist
  • NY Criminal Defense Lawyer
  • Manhattan Family Law Specialist
  • Queens Immigration Lawyer
  • Staten Island Civil Rights Advocate
  • Manhattan Personal Injury Prevention
  • Bronx Intellectual Property Attorney

Copyright © 2026 164news.com.

Powered by PressBook Dark WordPress theme