AP Logo

Menu

SECTIONS

Iran war Russia-Ukraine war Español China Asia Pacific Latin America Europe Africa

TOP STORIES

Newsletters

The Afternoon Wire\ \ Our personalization engine sends you what you may have missed throughout the day. The Morning Wire\ \ Our flagship newsletter breaks down the biggest headlines of the day.

See All Newsletters

SECTIONS

Immigration Weather Education Transportation Abortion LGBTQ+ Notable Deaths America at 250

TOP STORIES

Newsletters

The Afternoon Wire\ \ Our personalization engine sends you what you may have missed throughout the day.

See All Newsletters

AP QUIZZES

Weekly News Quiz

SECTIONS

2026 Elections Election Results Election calendar White House Congress Supreme Court The latest AP-NORC polls

TOP STORIES

Newsletters

Ground Game\ \ Your guide to the biggest stories in politics, policy and U.S. elections.

See All Newsletters

SECTIONS

NFL MLB NBA WNBA Soccer Tennis Golf NHL

TOP STORIES

Newsletters

The Sports Wire\ \ Get AP’s in-depth sports reporting tailored to your interests.

See All Newsletters

SECTIONS

Movies Fashion Television Celebrity Interviews Music Books

TOP STORIES

Newsletters

AP Entertainment Wire\ \ AP’s entertainment, media, fashion and culture coverage, personalized to your interests.

See All Newsletters

SECTIONS

Tariffs Inflation Financial Markets Financial Wellness Technology

TOP STORIES

SECTIONS

Space Animals The Ancient World Climate Medicine

TOP STORIES

TOP STORIES

TOP STORIES

SECTIONS

Trending Better health At home Working well For the climate Eating well

TOP STORIES

SECTIONS

Photo Essays

TOP STORIES

Newsletters

The World in Pictures\ \ Get special access to AP's most compelling images from our global team of photojournalists.

See All Newsletters

SECTIONS

Indigenous peoples and climate Climate Questions Climate Migration India Focus

TOP STORIES

TOP STORIES

SECTIONS

Artificial Intelligence Social Media

TOP STORIES

SECTIONS

Food & Recipes Gardening Fashion Homes Travel Pets

TOP STORIES

TOP STORIES

Newsletters

World of Faith\ \ Comprehensive global coverage of how religion shapes our world.

See All Newsletters

SECTIONS

Política de EEUU Deportes

TOP STORIES

MORE

Sign in

Search Query Submit Search

Show Search Menu

Submit Search

Mitch McConnell discharged

Tennessee primary results

Iran war

Sydney Towle dies

Hall of Fame game

AP Logo

Menu

SECTIONS

Iran war Russia-Ukraine war Español China Asia Pacific Latin America Europe Africa

TOP STORIES

Newsletters

The Afternoon Wire\ \ Our personalization engine sends you what you may have missed throughout the day. The Morning Wire\ \ Our flagship newsletter breaks down the biggest headlines of the day.

See All Newsletters

SECTIONS

Immigration Weather Education Transportation Abortion LGBTQ+ Notable Deaths America at 250

TOP STORIES

Newsletters

The Afternoon Wire\ \ Our personalization engine sends you what you may have missed throughout the day.

See All Newsletters

AP QUIZZES

Weekly News Quiz

SECTIONS

2026 Elections Election Results Election calendar White House Congress Supreme Court The latest AP-NORC polls

TOP STORIES

Newsletters

Ground Game\ \ Your guide to the biggest stories in politics, policy and U.S. elections.

See All Newsletters

SECTIONS

NFL MLB NBA WNBA Soccer Tennis Golf NHL

TOP STORIES

Newsletters

The Sports Wire\ \ Get AP’s in-depth sports reporting tailored to your interests.

See All Newsletters

SECTIONS

Movies Fashion Television Celebrity Interviews Music Books

TOP STORIES

Newsletters

AP Entertainment Wire\ \ AP’s entertainment, media, fashion and culture coverage, personalized to your interests.

See All Newsletters

SECTIONS

Tariffs Inflation Financial Markets Financial Wellness Technology

TOP STORIES

SECTIONS

Space Animals The Ancient World Climate Medicine

TOP STORIES

TOP STORIES

TOP STORIES

SECTIONS

Trending Better health At home Working well For the climate Eating well

TOP STORIES

SECTIONS

Photo Essays

TOP STORIES

Newsletters

The World in Pictures\ \ Get special access to AP's most compelling images from our global team of photojournalists.

See All Newsletters

SECTIONS

Indigenous peoples and climate Climate Questions Climate Migration India Focus

TOP STORIES

TOP STORIES

SECTIONS

Artificial Intelligence Social Media

TOP STORIES

SECTIONS

Food & Recipes Gardening Fashion Homes Travel Pets

TOP STORIES

TOP STORIES

Newsletters

World of Faith\ \ Comprehensive global coverage of how religion shapes our world.

See All Newsletters

SECTIONS

Política de EEUU Deportes

TOP STORIES

MORE

Sign in

Search Query Submit Search

Show Search Menu

Submit Search

Business

Meta says its AI model hacked another company, adding to worries about bots going rogue

A Meta logo is shown on a video screen at LlamaCon 2025, an AI developer conference, in Menlo Park, Calif., April 29, 2025. (AP Photo/Jeff Chiu, File)

A Meta logo is shown on a video screen at LlamaCon 2025, an AI developer conference, in Menlo Park, Calif., April 29, 2025. (AP Photo/Jeff Chiu, File)

Read More

By BARBARA ORTUTAY

Updated [hour]:[minute] [AMPM] [timezone], [monthFull] [day], [year] Updated 4:38 PM EDT, August 6, 2026

Leer en español

Add AP News on Google Add AP News as your preferred source to see more of our stories on Google. Share

Comments

Share

Link copied

Meta said Thursday that one of its artificial intelligence models accessed the internet on its own and hacked another company, the latest in a series of disclosures about AI models going rogue.

In recent weeks OpenAI and Anthropic also have described instances of AI models going beyond humans’ instructions to access the web and find ways around other companies’ digital security.

Meta said in a statement that a “misconfiguration” during cybersecurity testing by Irregular, an independent company hired by Meta, inadvertently allowed one of its models to access the internet.

“The model subsequently exploited a security vulnerability in a third-party service, in a manner similar to previously-reported instances with other companies,” the company said. Meta said it is investigating the incident and will issue a report when that’s complete.

The disclosure has added to worries about AI models acting autonomously.

Separately this week, the United Kingdom’s AI Security Institute announced it had found “unsanctioned agent behavior” during cyber testing. In one case, an agent created fake online identities to pressure a person to approve use of malicious code.

“On investigation, we found that some of the agents being tested had engaged in sustained, potentially harmful activity directed at real people and organizations,” AISI said Tuesday. “We declared a security incident and, within roughly one hour of discovery, had contained it and begun a full investigation.”

Related Stories

Anthropic says its AI models hacked 3 organizations during testing

Anthropic says its AI models hacked 3 organizations during testing

1 MIN READ

26 Meta employees sue, alleging AI-driven layoff picks hit workers on medical and parental leave

26 Meta employees sue, alleging AI-driven layoff picks hit workers on medical and parental leave

3 MIN READ

51

OpenAI says its AI technology acted on its own in an ‘unprecedented’ hack of another company

OpenAI says its AI technology acted on its own in an ‘unprecedented’ hack of another company

3 MIN READ

51

During the agency’s testing, Anthropic and OpenAI models took “autonomous, unsanctioned action” on the internet. Some guardrails to prevent misuse had been disabled, the agency said.

“As was standard in our cyber testing, we had intentionally permitted internet access, and model-provider cyber classifiers were deliberately disabled — conditions that do not reflect how frontier models are made available to the public,” AISI said. “We do this to best assess the maximum capability of models.”

Anthropic said it is “grateful” for AISI’s work and added that it underscores the need for a broader conversation about how to safely evaluate AI agents as their capabilities grow.

OpenAI said the AISI incidents took place “in testing environments with reduced safeguards, under conditions that do not reflect ordinary use.” It added it will continue working with others across the industry to “strengthen shared practices for conducting evaluations safely as models become more capable.”

Sign up for Morning Wire: Our flagship newsletter breaks down the biggest headlines of the day.

Email address

Sign up

By checking this box, you agree to AP's Terms of Use and acknowledge that AP may collect and use your data pursuant to our Privacy Policy.

The first company to disclose a hack late last month, OpenAI said it had tasked the AI models involved with pursuing “advanced exploitation using complex attack paths” to test cyber capabilities, but the technology went to unexpected lengths. It apparently decided on its own to target Hugging Face, a well-known AI development hub and marketplace, to obtain information it needed to carry out a task.

A spokesperson for Irregular, the San Francisco-based AI security company, said the Meta episode involves a test-environment issue that was disclosed last week by Anthropic.

Irregular said it’s writing a paper to share “best practices for containment” to prevent such incidents in the future and securely run cyber tests.

BARBARA ORTUTAY

Ortutay writes about social media and the internet for The Associated Press.

mailto

Read Original at AP News