Jauhn Etienne Villaruel

Article (news)

Agence France-Presse

<div><span class="fr-img-caption fr-fic fr-dib" style="width:960px;display:block" width="960"><span class="fr-img-wrap"><img src="https://od2-image-api.abs-cbn.com/prod/editorImage/1785458610959431302153_756477769916599_5611869104617254758_n (2).jpg" style="width:960px" class="fr-fic fr-dib" width="960"/><span class="fr-inner" style="text-align:left">The San Francisco-based company announced that the US Department of Commerce had ordered it to suspend Mythos 5 and Fable 5 for &quot;national security&quot; reasons, without providing further details. <em>📷༘ Anthropic/Facebook</em><br/></span></span></span></div><p>SAN FRANCISCO, United States — Anthropic&#x27;s artificial intelligence (AI) models &quot;gained unauthorized access&quot; to three outside organizations during testing that was supposed to keep them away from &quot;real-world&quot; systems, the company said on Thursday.</p><div><a href="https://www.abs-cbn.com/news/technology/2026/6/5/anthropic-calls-for-pause-of-global-ai-development-0737">Anthropic calls for pause of global AI development</a></div><p>The announcement comes just days after rival OpenAI first revealed that its models improperly accessed the internet and went rogue during security testing.</p><p>Anthropic evaluated more than 141,000 &quot;evaluation runs&quot; and found that three different versions of its model, known as Claude, improperly accessed the systems of three unnamed organizations.</p><p>Unlike the incident involving OpenAI&#x27;s technology, Anthropic&#x27;s models had access to the internet &quot;due to a misunderstanding between us and our evaluation partner,&quot; called Irregular, Anthropic said in a blog post.</p><p>Nonetheless, Claude used &quot;basic techniques, such as exploiting weak passwords and unauthenticated endpoints,&quot; the blog continued.</p><div><a href="https://www.abs-cbn.com/news/technology/2026/7/30/openai-says-rogue-ai-agent-attack-hit-other-companies-0820">OpenAI says rogue AI agent attack hit other companies</a></div><p>The models involved one of its most powerful ones known as Mythos 5, which has only been released to a limited number of approved partners.</p><p>Anthropic is working with Irregular to assess the situation, it said, and the company has contacted or attempted to contact all three impacted organizations.</p><p>On Tuesday, OpenAI confirmed that its models breached multiple companies.</p><p>It admitted last week that its models broke out of their confined environment during testing, connected to the internet, and infiltrated Hugging Face, a site where developers store and share their code.</p><p>Days later, OpenAI said it found three additional incidents.</p><p>OpenAI CEO Sam Altman said on a podcast this week that the company had &quot;paused&quot; its own testing after the incident while it improved the security around its &quot;sandboxing,&quot; which is the process of isolating software in a controlled environment for testing.</p><p>The incident also triggered a petition signed by over 1,000 employees at cutting-edge AI companies calling on the US government to help slow the release of the most advanced AI models.</p><div><a href="https://www.abs-cbn.com/news/technology/2026/7/29/ai-industry-slowdown-may-be-needed-after-security-scare-leaders-warn-0914">AI industry slowdown may be needed after security scare, leaders warn</a></div><p>Anthropic CEO Dario Amodei was among those who signed the petition.</p><p>Titled &quot;Pacing the Frontier,&quot; the petition requests &quot;that the US government support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development.&quot;</p><p>Altman did not sign the petition, but during the podcast, he suggested the tech industry might need to slow down development of advanced models.</p><p>&quot;We may have to pace the rate of AI development to give ourselves enough time for society to harden around some of these new capability levels,&quot; Altman said.</p><div><br/></div><div><strong>© Agence France-Presse</strong></div><div><strong><br/></strong></div><div><strong>RELATED VIDEOS</strong></div><div><span class="fr-video fr-deletable fr-fvc fr-dvb fr-draggable" contenteditable="false" draggable="true"><iframe width="640" height="360" src="https://www.youtube.com/embed/ZF31IdqtUI8?&amp;wmode=opaque&amp;rel=0" frameBorder="0" allowfullscreen="" class="fr-draggable"></iframe></span><br/></div>

<p>Anthropic's artificial intelligence (AI) models "gained unauthorized access" to three outside organizations during testing that was supposed to keep them away from "real-world" systems, the company said on Thursday.</p>

Anthropic's artificial intelligence (AI) models "gained unauthorized access" to three outside organizations during testing that was supposed to keep them away from "real-world" systems, the company said on Thursday.

Anthropic's models gained unauthorized 'real-world' access during testing

News

Entertainment

Lifestyle

Sports

Metro.Style

More

BINI

Sudoku

Word of the Day

Lotto

Horoscope

Technology

Anthropic's models gained unauthorized 'real-world' access during testing

Agence France-Presse

Published July 31, 2026 08:43 AM PHT

Read More:

San Francisco

United States

us

tech

ai

anthropic

Agence France-Presse

ABS-CBN Newsletter

Your go-to source for the latest news, entertainment highlights, exclusive interviews, and updates on your favorite Kapamilya stars.

ABS-CBN Newsletter

Your go-to source for the latest news, entertainment highlights, exclusive interviews, and updates on your favorite Kapamilya stars.

Sign Up

I have read and agree to the Terms and Privacy Policy.

ABS-CBN is the leading media and entertainment company in the Philippines, offering quality content across TV, radio, digital, and film. Committed to public service and promoting Filipino values, ABS-CBN continues to inspire and connect audiences worldwide.

Corporate

Investors

Governance

International Distribution

Corporate

Governance

Investors

International Distribution

NPC Seal of Registration

Privacy Policy

AI Policy

Terms of Service

Advertise with Us

© 2026 ABS-CBN Corporation. All Rights Reserved.

NPC Seal of Registration

Privacy Policy

Terms of Service

AI Policy

Advertise with Us

© 2026 ABS-CBN Corporation. All Rights Reserved.

NPC Seal of Registration

Privacy Policy

Terms of Service

AI Policy

Advertise with Us

© 2026 ABS-CBN Corporation. All Rights Reserved.