Get all your news in one place.

100's of premium titles.
One app.

Get all your news in one place.

100's of premium titles. One news app.

THE TRUTH ABOUT: Irregular - The Untold Story

AI lab's safety systems are falling behind

Recent incidents involving OpenAI, Anthropic, and Meta show what happens when increasingly capable AI agents are tested in flawed environments. A new assessment finds leading labs are better at spotting risky behavior than reliably stopping it.

AI Models Went Rogue at OpenAI, Anthropic and Meta. One Small Israeli Startup Reportedly Links the Incidents

AI Models Went Rogue at OpenAI, Anthropic and Meta. One Small Israeli Startup Reportedly Links the Incidents

The company, Irregular, is a three-year-old Tel Aviv startup that provides specialized testing environments designed to probe the capabilities and potential security risks of increasingly powerful AI systems.

International Business Times

Meta’s AI hacked another company and now the industry’s ‘under control’ narrative takes another hit

Meta’s AI hacked another company and now the industry’s ‘under control’ narrative takes another hit

A pattern that's getting harder to ignore.

We Got This Covered

From analysis to the latest developments in health,
read the most diverse news in one place.

inkl

Another AI Hacking: Meta Model Slipped Into Another Company's Systems During Testing

Meta has confirmed that one of its AI models hacked into another company's systems during cybersecurity testing after a configuration error inadvertently gave it internet access, adding to similar incidents recently disclosed by OpenAI and Anthropic.

International Business Times

Meta Admits Its AI Model Hacked Another Company: 'Rogue AI' Is Becoming a Pattern

Meta's AI model accessed external systems during testing, raising concerns over AI cyber capabilities and vulnerabilities, similar to recent incidents at OpenAI and Anthropic.

International Business Times UK

AI Models Go Rogue Again: OpenAI and Anthropic Models Attempt Unauthorized Hacks & Communication

New testing revealed that artificial intelligence agents from OpenAI and Anthropic carried out unauthorized hacking attempts as concerns about their ability to go rogue escalate.

International Business Times

Meta says its AI has gone rogue and hacked other companies

Facebook and Instagram owner joins ChatGPT’s OpenAI and Claude’s Anthropic in declaring major cyber attacking incidents

The Independent UK

inkl

Read news from The Economist, FT, and more,
with one subscription

Meta AI model hacks another company during testing

Meta AI model hacks another company during testing

In a recent security testing incident, Meta disclosed that an AI model infiltrated the company, echoing breaches at Anthropic and OpenAI. This series of events amplifies concerns regarding the interplay between advanced AI technologies and cybersecurity threats. In response, U.S. government initiatives aimed at enhancing AI safety are set to…

The Economic Times

Meta Joins OpenAI and Anthropic in AI Cybersecurity Scare After Model Hacks Third Party: 'We Are Currently Investigating…'

Meta Joins OpenAI and Anthropic in AI Cybersecurity Scare After Model Hacks Third Party: 'We Are Currently Investigating…'

Meta Platforms, Inc. (NASDAQ:META) has joined OpenAI and Anthropic in facing fresh AI safety concerns after one of its models gained unintended internet access during cybersecurity testing and exploited a security vulnerability involving another company.

OpenAI, Anthropic model tests reveal more ‘unsanctioned’ actions

OpenAI, Anthropic model tests reveal more ‘unsanctioned’ actions

AI models from OpenAI and Anthropic demonstrated harmful actions during safety tests. These systems engaged in hacking and attempted code injection, surprising researchers. The UK's AI Security Institute observed these "unsanctioned" and autonomous activities. Both companies are investigating these incidents and their implications for AI safety. This highlights the need…

The Economic Times

Anthropic sees OpenAI cybersecurity disaster and says 'hold my beer,' reveals it accidentally hacked 3 companies in as many months without noticing

Anthropic sees OpenAI cybersecurity disaster and says 'hold my beer,' reveals it accidentally hacked 3 companies in as many months without noticing

I'm sure this is fine.

Number of AI chatbots ignoring human instructions increasing, study says

Number of AI chatbots ignoring human instructions increasing, study says

Exclusive: Research finds sharp rise in models evading safeguards and destroying emails without permission

The Guardian - UK

'No one asked them to': Security experts warn malicious AI agents can team up to launch cyberattacks

'No one asked them to': Security experts warn malicious AI agents can team up to launch cyberattacks

AI agents which perform normal office tasks can also autonomously exploit systems, bypass protections, and exfiltrate sensitive data inside simulated networks.

‘Exploit every vulnerability’: rogue AI agents published passwords and overrode anti-virus software

‘Exploit every vulnerability’: rogue AI agents published passwords and overrode anti-virus software

Exclusive: Lab tests discover ‘new form of insider risk’ with artificial intelligence agents engaging in autonomous, even ‘aggressive’ behaviours

The Guardian - UK

Get all your
news in one place.

100's of premium titles.
One news app.

Hacking is about to get very cheap thanks to AI

As AI agents become cheap, tireless attackers, the old assumptions behind corporate cybersecurity are breaking down.

1

People

From analysis to the latest developments in health,
read the most diverse news in one place.

inkl

Organisations

inkl

Read news from The Economist, FT, and more,
with one subscription

Topics

Artificial Intelligence

Places

415,555 stories

148,634 stories

Fourteen days free

Download the app

One app. One membership.
100+ trusted global sources.

Download on the AppStore

Get it on Google Play