Illustration of US Treasury Secretary warning bank executives about AI cyberattack risks from Anthropic's Claude Mythos.
Illustration of US Treasury Secretary warning bank executives about AI cyberattack risks from Anthropic's Claude Mythos.
Imagen generada por IA

Anthropic anuncia IA Claude Mythos capaz de explotar vulnerabilidades de software

Imagen generada por IA

La startup Anthropic ha revelado Claude Mythos, un modelo de IA con capacidad para identificar y explotar brechas de ciberseguridad en sistemas operativos y navegadores. Ante el riesgo de uso por ciberdelincuentes, no lo liberará públicamente y lo restringe a una coalición de empresas tecnológicas. El secretario del Tesoro de EE UU ha convocado a bancos para alertar sobre estos peligros.

Anthropic, startup rival de OpenAI, anunció esta semana Claude Mythos, un modelo de IA general que destaca por su habilidad en codificación. Según la compañía, supera a la mayoría de humanos en encontrar y explotar vulnerabilidades de software, habiendo detectado miles de fallos en grandes sistemas operativos y navegadores web, incluido un bug de 27 años en OpenBSD.

"Los modelos de IA han alcanzado un nivel de capacidad de codificación que les permite superar a todos, salvo a los humanos más habilidosos", afirma Anthropic en su comunicado. Logan Graham, del equipo de pruebas, lo califica de “punto de inflexión” para la industria. Por temor a ciberdelincuentes, Anthropic no lo libera al público y lo despliega vía Project Glasswing, coalición de 40 organizaciones como Apple, Amazon, Microsoft, Google y Linux Foundation, para parchear brechas.

El debate escaló cuando el secretario del Tesoro, Scott Bessent, convocó esta semana a directivos de Bank of America, Citigroup, Goldman Sachs, Morgan Stanley y Wells Fargo, más Jay Powell de la Fed. Fuentes citadas por Financial Times indican que alertó sobre riesgos de IA en ciberataques. Jamie Dimon de JPMorgan, ausente, había advertido previamente de estos peligros agravados por IA.

No obstante, el analista Gary Marcus cuestiona el anuncio como “inflado” por marketing, señalando que pruebas en Firefox desactivaron sandboxing y que modelos abiertos ya logran similares hazañas.

Artículos relacionados

Illustration of Anthropic restricting Claude Mythos AI and launching Project Glasswing consortium with tech giants to address cybersecurity vulnerabilities.
Imagen generada por IA

Anthropic restricts Claude Mythos AI release and launches Project Glasswing over cybersecurity risks

Reportado por IA Imagen generada por IA

Anthropic has limited access to its Claude Mythos Preview AI model due to its superior ability to detect and exploit software vulnerabilities, while launching Project Glasswing—a consortium with over 45 tech firms including Apple, Google, and Microsoft—to collaboratively patch flaws and bolster defenses. The announcement follows recent data leaks at the firm.

Germany's financial regulator BaFin has warned banks about risks from Anthropic's Claude Mythos AI model, following US Treasury alerts. The model autonomously detects IT vulnerabilities at scale, potentially accelerating cyberattacks. US banks are testing it amid restrictions.

Reportado por IA

The UK government’s AI Security Institute has released an evaluation of Anthropic's Mythos Preview AI model, confirming its strong performance in multistep cyber infiltration challenges. Mythos became the first model to fully complete a demanding 32-step network attack simulation known as 'The Last Ones.' The institute cautions that real-world defenses may limit such automated threats.

Anthropic has launched Claude Opus 4.7, a new AI model designed to assist developers with complex coding tasks. The company emphasized its improved instruction-following and memory capabilities. This release follows the earlier announcement of the more advanced Claude Mythos Preview.

Reportado por IA

Anthropic's Claude Cowork AI tool has caused a sharp decline in stocks of Infosys, TCS, and other SaaS companies. These firms lost hundreds of billions of dollars in market value. The trigger is the rise of AI.

Anthropic's CEO Dario Amodei stated that the company will not comply with the Pentagon's request to remove safeguards from its AI models, despite threats of exclusion from defense systems. The dispute centers on preventing the AI's use in autonomous weapons and domestic surveillance. The firm, which has a $200 million contract with the Department of Defense, emphasizes its commitment to ethical AI use.

Reportado por IA

Anthropic has announced that its AI chatbot Claude will remain free of advertisements, contrasting sharply with rival OpenAI's recent decision to test ads in ChatGPT. The company launched a Super Bowl ad campaign mocking AI assistants that interrupt conversations with product pitches. This move highlights growing tensions in the competitive AI landscape.

 

 

 

Este sitio web utiliza cookies

Utilizamos cookies para análisis con el fin de mejorar nuestro sitio. Lee nuestra política de privacidad para más información.
Rechazar