Artificial intelligence theft? Here’s why Anthropic is accusing the Chinese company Alibaba
Anthropic has accused the Chinese group Alibaba (which the Pentagon considers close to Beijing’s armed forces) of illegally extracting information from Claude: this practice is known as “distillation,” which Chinese AI companies often use to keep the costs of their models low. All the details.
The battle between Musk and Altman ends up in court, but the conflict over the future of AI remains unresolved.
The lawsuit against OpenAI closes without addressing the substance of Elon Musk’s accusations, leaving unresolved issues regarding the shift to profitability and the governance of artificial intelligence, while global competition between models intensifies and the risks of still fragmented governance remain high. A conversation between Stefano Feltri and intellectual property lawyer Laura Turini for Appunti.
Surprise praise for the lobbyist who finally speaks bluntly to major newspapers and politicians
Why did Veltroni’s predictable and at times trivial interview with AI Claude in Corriere della Sera startle Diego Ciulli from Google? The letter from Claudio Trezzano.
The latest Claude from Anthropic has uncovered thousands of vulnerabilities in software.
Anthropic restricts access to the new Claude Mythos Preview model after it discovered thousands of software vulnerabilities and launches an alliance with over 40 companies, including Apple, Amazon, Microsoft, Google, Cisco, and Broadcom, to strengthen cybersecurity. All the details.
Does a human error leave Claude from Anthropic in the lurch?
A major setback for Claude, the AI that many users disappointed by the pro-government stance of rival OpenAI are flocking to: a lapse in attention exposed the source code, leaving it vulnerable to rival software companies as well as hackers. And this is not the first leak to hit Anthropic after the “great refusal” to the Pentagon…
How Anthropic, Palantir, and Starlink Helped the US and Israel Attack Iran
In the joint US-Israel raid in Iran, Anthropic’s AI was reportedly used for intelligence analysis, target identification, and battle simulations, integrated into the Maven Smart System developed in collaboration with Palantir. Meanwhile, Starshield, SpaceX’s military variant of Starlink, provided secure and jam-resistant satellite communications during the operation.
