• Home
  • About US
  • Contact Us
  • Privacy Policy
  • Terms of Use

Archives

  • November 2026
  • October 2026
  • September 2026
  • August 2026
  • July 2026
  • June 2026
  • May 2026
  • April 2026
  • March 2026
  • February 2026
  • January 2026
  • December 2025
  • November 2025
  • October 2025
  • September 2025
  • August 2025
  • July 2025
  • June 2025
  • May 2025
  • April 2025
  • March 2025
  • February 2025
  • January 2025
  • December 2024
  • November 2024
  • October 2024
  • September 2024
  • August 2024
  • July 2024
  • June 2024
  • May 2024
  • April 2024
  • March 2024
  • February 2024
  • January 2024
  • December 2023
  • November 2023
  • October 2023
  • September 2023
  • August 2023
  • July 2023
  • June 2023
  • May 2023
  • April 2023
  • March 2023
  • January 2023
  • December 2022
  • November 2022
  • October 2022
  • September 2022
  • August 2022
  • July 2022
  • June 2022
  • May 2022
  • March 2022
  • February 2022
  • January 2022
  • November 2019

Categories

  • Auto
  • Business
  • Climate & Earth
  • Coronavirus
  • Crypto
  • Entertainment
  • Happiness Basket
  • India
  • Learn | Unlearn | Relearn
  • Lifestyle
  • Politics
  • Research Discoveries
  • Science & Technology
  • Sports
  • Trending
  • Video
  • World
  • Home
  • About US
  • Contact Us
  • Privacy Policy
  • Terms of Use
Read Selective
  • World
  • Politics
  • India
  • Business
  • Entertainment
  • Lifestyle
  • Auto
  • Crypto
  • Coronavirus
  • Happiness Basket
  • Research Discoveries
  • Science & Technology

OpenAI, Anthropic AI Agents Breach Security Again, Create Fake Profiles For Testing

  • August 5, 2026

Anthropic and OpenAI’s agents engaged in unauthorised actions during security evaluations the government organization conducted to assess the models’ capabilities.

Anthropic said it was working closely with AISI to obtain more details and conduct its own investigation.

An AI agent was caught creating fake online identities to gain unauthorised access to secure systems during tests of models from OpenAI and Anthropic which revealed a series of new breaches, Britain’s AI Security Institute (AISI) disclosed on Tuesday.

The institute said agents powered by Anthropic’s Mythos 5 and OpenAI’s GPT-5.6-Sol engaged in unauthorised actions during security evaluations the government organization conducted to assess the models’ capabilities.

“Some of the agents being tested had engaged in sustained, potentially harmful activity directed at real people and organisations,” AISI said in a blog post.

The report underscores the lax state of safeguards around the process of testing agents, which AI companies are simultaneously marketing as the future of business.

AISI, which receives access to advanced AI models under voluntary agreements from major labs, put the agents through a fictional cybersecurity scenario to test their capabilities.

It ran the challenge 122 times, and identified 19 unsanctioned actions across a total of 10 test runs. Anthropic’s agent was behind 17 of the actions, and OpenAI’s agent the remaining two.

The most egregious action involved an agent writing malicious code and creating fake online identities in an attempt to get a human to approve the code, AISI said, adding that no real-world harm was found as a result of any of the breaches.

While AISI did not say which agent was behind the fake identities, the breach did not match either of the two cases that OpenAI self-disclosed.

Andrew Yoon, a researcher at CivAI, a California non-profit that examines AI capabilities and dangers, said it appeared that Anthropic’s agent was responsible.

“The fact that Mythos engaged in such deceptive actions, with apparent awareness that it was targeting a real person, suggests that Anthropic does not have as good a handle on their models as they think,” Yoon said.

In a statement on X, Anthropic said it was working closely with AISI to obtain more details and conduct its own investigation.

OpenAI shared details in a company blog post, noting that both of its agent’s unapproved actions involved accessing the internet in ways that were forbidden by the prompt.

“We are committed to working across the industry to strengthen shared practices for conducting high-risk evaluations safely, including convening stakeholders such as national AI institutes, independent evaluators, other AI labs, and other groups in the coming weeks,” OpenAI said.

Source : https://www.ndtv.com/artificial-intelligence/openai-anthropic-ai-agents-breach-security-again-create-fake-profiles-for-testing-11866864?pfrom=home-ndtv_topstories

Previous Article
  • Science & Technology

The Sound Of Disease: Your Voice Could Become Medicine’s Next ‘Vital Sign’

  • July 30, 2026
View Post
Next Article
  • Science & Technology

NASA Plans Permanent Base On Moon, Paving The Way For Future Mars Missions

  • August 6, 2026
View Post
You May Also Like
View Post
  • Science & Technology

Anthropic Shares Plunge Over 9% After Record IPO Debut As AI Stocks Slide; Nvidia, Meta, AMD & Intel Face Selling Pressure

  • November 21, 2026
View Post
  • Science & Technology

GTA 6’s Bengaluru Connection: How 1,600 Developers Helped Build The World’s Biggest Game

  • November 19, 2026
View Post
  • Science & Technology

Surge in software patches likely behind recent tech outages, experts say

  • November 19, 2026
View Post
  • Science & Technology

‘Many Services Are Back Up And Running’: Apple Confirms Outage, Reveals What Triggered Disruption

  • November 18, 2026
View Post
  • Science & Technology

US Rocked by Another Mass Outage As Banking Giants Go Down, Customers Continue Facing Issues

  • November 17, 2026
View Post
  • Science & Technology

Microsoft’s Global Meltdown: How India’s Govt Digital Backbone Stayed Up & Running

  • November 15, 2026
View Post
  • Science & Technology

Microsoft restores service to Outlook and Teams after global outage

  • November 14, 2026
View Post
  • Science & Technology

Defence Pensioners Generate Over 20.48 Lakh Digital Life Certificates Under Nationwide DLC Campaign 5.0 Through Aadhaar Face Authentication

  • November 13, 2026

Recent Posts

  • India blind cricket team boycotts ceremony after Pak awarded rain-hit Asia Cup final
  • Students throw firecrackers at Italy’s education ministry as protests escalate
  • Rs 40 lakh in 8 bundles: Gujarat trader duped with ‘Children Bank of India’ notes
  • Trump praises India’s low interest rates, slams US central bank for being political
  • Locust Brain Sniffs Out Bacteria Hiding From Antibiotics, Study Finds
Categories
  • Auto (43)
  • Business (485)
  • Climate & Earth (17)
  • Coronavirus (17)
  • Crypto (26)
  • Entertainment (903)
  • Happiness Basket (7)
  • India (6,419)
  • Learn | Unlearn | Relearn (163)
  • Lifestyle (336)
  • Politics (102)
  • Research Discoveries (461)
  • Science & Technology (474)
  • Sports (1,064)
  • Trending (1,821)
  • Video (1)
  • World (10,373)
Read Selective

For Feedbacks, Advertisements or Any Other Concerns mail us at info@readselective.com

Pages
  • Home
  • About US
  • Contact Us
  • Privacy Policy
  • Terms of Use
Categories
  • Auto
  • Business
  • Climate & Earth
  • Coronavirus
  • Crypto
  • Entertainment
  • Happiness Basket
  • India
  • Learn | Unlearn | Relearn
  • Lifestyle
  • Politics
  • Research Discoveries
  • Science & Technology
  • Sports
  • Trending
  • Video
  • World
© 2024 Read Selective | Developed by SUGARA Technologies
  • Home
  • About US
  • Contact Us
  • Privacy Policy
  • Terms of Use

Input your search keywords and press Enter.

Go to mobile version