Dario Amodei, Anthropic’s boss, is sounding the alarm on his lab’s latest model. Mythos, which he says is too dangerous to make widely available, has reportedly found severe vulnerabilities in “every major operating system and web browser”—including one flaw that went undetected for 27 years. People have long warned that AI could be a menace for cyber-security. Is that moment here?

A world of “vibe hacking”, in which amateurs can use AI models to find flaws in software—and perhaps even write the “exploits” needed to crash them, hold them to ransom or even take control of them remotely—sounds terrifying. Shortly after Anthropic’s announcement Scott Bessent, America’s treasury secretary, hosted a meeting of bank bosses to discuss what ai-enabled hacking might mean for their businesses

Smart glasses are back — but this time powered by AI.
Warby Parker and Google are betting that generative AI will finally make wearable tech mainstream. Is this the beginning of the post-smartphone era?

China is accelerating the humanoid robotics race — but technological prowess does not guarantee market demand. The real question is not who can build them, but who will deploy them at scale.Click to watch an excerpt from the Chinese New Year show and see these robots in action.

Artificial intelligence is not just a speculative bubble: for Big Tech, the real challenges are now economic, strategic and regulatory.

This study examines the impact of generative AI-powered recommendations on potential tourists’ perceptions, focusing on the interplay between social presence and privacy concerns.

In destination marketing, generative AI can reduce the effectiveness of human-centered visuals due to privacy concerns—unless transparency and data-use justification are clearly communicated.

Meta Platforms Inc. Chief Executive Officer Mark Zuckerberg announced a major restructuring of the company’s artificial intelligence group, including a commitment to developing AI “superintelligence,” or systems that can complete tasks as well as or even better than humans.

Si votre stratégie IA se limite à gagner en productivité, vous êtes déjà en retard. L’IA n’est pas une machine à optimiser les proces

Christmas stockings may contain more surprises than usual this year, as children open presents that can talk back. Toymakers in China have declared 2025 the year of artificial intelligence (ai) and are producing robots and teddies that can teach, play and tell stories. Older children, meanwhile, are glued to viral ai videos and ai-enhanced games. At school, many are being taught with materials created with tools like Chatgpt. Some are even learning alongside chatbot-tutors.

A handpicked article read aloud from the latest issue of The Economist. The world’s most powerful new technology is in the hands of just a few men. While the Trump administration previously maintained a hands-off approach to ai, it appears America is finally waking up to the dangers.

As artificial intelligence becomes more powerful, its risks are no longer theoretical. The debate is shifting from innovation to control - and from opportunity to responsibility.

As the discreet wearable cameras become more popular, students are saying they feel constantly watched and harassed - and professors are reshaping their classrooms in response.

Everyone has accepted the reality that the second you step outside, you waive your right to privacy.

Modern artificial intelligence is seemingly unstoppable. But large language models have an undeniable blind spot: an incomplete understanding of physical reality and the laws of physics. Could “world models” be the solution—and how are AI labs building them?