Pentagon partners with tech firms to boost AI capabilities
The Pentagon has signed agreements with Nvidia, Microsoft, Amazon, and Reflection AI to enhance the deployment of advanced artificial intelligence tools on classified military networks for lawful operational use. These deals are part of the Department of Defense's strategy to diversify its AI vendors following a dispute with Anthropic over usage terms. The agreements aim to strengthen decision-making capabilities for over 1.3 million personnel using the secure enterprise platform GenAI.mil.
Anthropic's AI Model Raises Cybersecurity Concerns
Anthropic's Claude Mythos AI model has demonstrated the ability to identify and exploit vulnerabilities in financial software at unprecedented speeds, raising cybersecurity concerns among governments and the tech sector. The model can find zero-day bugs across all major operating systems and web browsers, including a 27-year-old flaw in OpenBSD. In response, a consortium of leading software providers has launched Project Glasswing to leverage the model for cybersecurity defense before it can be exploited by malicious actors.
Meta Acquires Assured Robot Intelligence to Boost AI Robotics
Meta Platforms Inc. has acquired Assured Robot Intelligence, a startup focused on developing AI models for robots, to enhance its humanoid technology initiatives. The acquisition, finalized on Friday, aims to leverage ARI's expertise in robotic intelligence to enable robots to understand and adapt to human behaviors in complex environments. Financial terms of the deal were not disclosed, but ARI's team, including co-founders Lerrel Pinto and Xiaolong Wang, will join Meta's Superintelligence Labs research division.
Minnesota enacts first AI-generated nude image ban
Minnesota has enacted the nation's first ban on apps that create fake nude images using artificial intelligence, with the state Senate passing the bill unanimously at 65-0. The legislation allows survivors to sue app owners for damages and empowers the state attorney general to impose fines of up to $500,000 per violation. Advocates, including Maye Quade, emphasized the law's role in protecting individuals from nonconsensual deepfakes, which have become increasingly prevalent.
