To win the « cybersecurity battle », the CIA will deploy « AI colleagues » alongside its agents

Last year, the Central Intelligence Agency (CIA) counted over 300 AI projects and, for the first time in its history, deployed artificial intelligence to generate a complete intelligence report. Over the next few years, the U.S. intelligence agency plans to implant artificial intelligence at the core of its work process, particularly by integrating AI agents who will work alongside human agents at the CIA.

New AI colleagues

The CIA plans to increase the number of AI agents within its ranks. At a conference dedicated to AI, Michael Ellis, Deputy Director of the CIA, defined these new type of workers as "colleagues" powered by artificial intelligence. Over the next few years, they will be particularly integrated into the workflow of its analysts and used for the collection and analysis of intelligence.

A tense situation

Of course, the integration of AI within the CIA will not be limited to writing tasks and data processing. The deputy director of the CIA explains without hesitation: «The battle for cybersecurity will be a battle of artificial intelligence, and whoever masters the best AI models will have considerable power».

On the one hand, it is a matter of a battle for attention. The CIA must be able to keep an eye on the models used. On the other hand, it is simply a battle of technical performance. As an example, Anthropic is currently disrupting the cybersecurity world with Claude Mythos. This AI has recently performed feats, while seeming to take decisions.

The US government and Anthropic have been openly in conflict because the startup refused to loosen restrictions regarding the use of its tools for internal surveillance. The deputy director of the CIA explained that the American intelligence service «cannot allow the caprices of a single company» to constrain its use of AI.

👉🏻 Follow real-time tech news: add 01net to your Google sources, and subscribe to our WhatsApp channel.

Source : Nextgov