All news
Politics
Business
Technology
Incidents
Sports
Weather
Science
Entertainment
World
Real Estate
AI News
"Anthropic acknowledges that clients used Claude to attempt to develop biological weapons. We are entering very dangerous territory"Photo by geralt on Pexels
AI News
USA
Investments

"Anthropic acknowledges that clients used Claude to attempt to develop biological weapons. We are entering very dangerous territory"

CNN Portugal13 September 2026 at 22:18

Anthropic publicly acknowledged that some of its clients used the Claude artificial intelligence model to attempt to develop biological weapons. This admission was made by the company founded by Dario Amodei, who warned that the sector is "entering very dangerous territory" without proper security precautions.

The statement comes in the context of a joint request made by three major names in the tech industry. Dario Amodei, Sam Altman (CEO of OpenAI), and Elon Musk signed a petition calling for a pause in artificial intelligence development until stricter security measures are implemented.

Commentator Miguel Portas commented on this petition and considered that Donald Trump's previous statements on the topic "blew up" any possibility of government-imposed regulation in the United States. According to Portas, Trump's positions made any attempt to impose government limits on AI development in the United States unviable.

Why this matters

This article directly affects Portuguese and European readers following the AI regulation debate, as it shows that attempts to impose security standards in the US were compromised by political statements. The revelation that AI tools were used to attempt to develop biological weapons underscores the urgency of security issues that potentially affect all technology users.

About this summary

This is our short summary of a report published by CNN Portugal on 13 September 2026 at 22:18; the full text stays with the publisher. In our feed it sits under AI News. We currently carry 684 items in that section.

All items in this section →
Source

Related articles

AI News

China asks for "security barrier" for AI in face of internal advances and warnings in the USA

The Chinese minister of State Security, Chen Yixin, called for the creation of a "security barrier" around artificial intelligence, warning about the growing political, social and technological risks associated with this technology, at a time when both Beijing and Washington are expressing growing concerns about the dangers of AI.

RTP Notícias14/09/26, 05:30
AI News

AI Alerts "Bring to Mind a Classic Phrase of Portuguese Swagger: 'Hold Me Back or I'll Kill Them'"

Mário Figueiredo, a researcher at the Instituto de Telecomunicações, analyzes the warnings made by tech leaders such as Elon Musk and others about the dangers of Artificial Intelligence, comparing them to a popular Portuguese expression of exaggeration. The researcher considers that, although debating AI risks is legitimate, many of these warnings are overstated and serve more to capture media attention than to present concrete solutions to the identified problems.

CNN Portugal13/09/26, 23:08
AI News

King Charles will gather AI company leaders amid calls for technology slowdown

King Charles III of the United Kingdom will hold a meeting this week, in Scotland, with executives from major artificial intelligence companies, including Nvidia, Google, DeepMind, OpenAI, and Anthropic, to discuss responsible development of the technology. The meeting comes amid growing international calls for a slowdown in AI advancement, with experts warning of risks of global catastrophe. The event at Dumfries House, headquarters of the King's Foundation, also includes the participation of government authorities.

oglobo13/09/26, 23:04
AI News

Point of no return

An investigator from Anthropic, the company responsible for the Claude AI model, came forward to affirm that "AI could kill us all by the end of the decade," generating widespread discomfort. The article analyzes this declaration as a "point of no return" moment in the discussion about existential risks of artificial intelligence.

Jornal de Negócios13/09/26, 23:00