
President Donald Trump has delayed signing an executive order intended to evaluate AI models for security before their release. The order, which would involve the Office of the National Cyber Director, aims to address security concerns related to AI models like Anthropic’s Mythos and OpenAI’s GPT-5.5 Cyber. Trump expressed dissatisfaction with the language of the order, particularly a clause requiring AI companies to share models with the government before launch. The delay highlights the challenge of balancing AI innovation with security oversight.
Read original
© TechCrunch AIIn a candid discussion, Google Cloud's COO Francis de Souza emphasized the critical need for integrating security into AI strategies from the outset. He highlighted the risks of 'shadow AI' and the necessity for a consistent security posture across multiple cloud environments. Despite Google's commitment to a multicloud approach, recent incidents involving unauthorized API calls to Gemini models reveal vulnerabilities in their system. These challenges underscore the urgency of developing AI-native defenses and the ongoing struggle to keep pace with rapidly evolving threats. The conversation reflects the broader industry challenge of aligning security practices with the fast-paced evolution of AI technologies.
© TechCrunch AIAmazon's Bee wearable is an AI-powered wrist gadget designed to act as a personal assistant by recording, transcribing, and summarizing conversations. While it shows potential in professional settings by helping users keep track of meetings and discussions, its extensive data collection and cloud storage raise privacy concerns. The device requires significant mobile permissions and stores data in the cloud, which might deter privacy-conscious users. Despite its promise, Bee's current iteration may be too invasive for personal use, though it could evolve into a valuable tool for professionals with further development.
© TechCrunch AIFerrari is leveraging IBM's AI technology to transform its fan engagement strategy, focusing on storytelling and personalization. The partnership aims to make the Ferrari fan app more interactive and engaging, with features like AI-written race summaries and an AI companion for fan queries. This move reflects a broader trend in Formula One, where teams are using advanced data analytics to deepen fan connections. By tailoring content to individual preferences, Ferrari hopes to build lasting loyalty among its diverse and growing fanbase.
© Matt WolfeGoogle announced at I/O that it is replacing traditional search with an AI-driven engine, prioritizing AI-generated overviews.
© The Verge AIHackers are increasingly exploiting the 'personalities' of AI chatbots, using conversational tactics rather than technical skills to bypass safety protocols. This new wave of attacks involves manipulating chatbots through persuasive dialogue, revealing a vulnerability in AI systems that rely on human-like interactions. Companies have patched obvious loopholes, but the challenge remains in balancing useful conversation with security. As AI systems become more integrated into daily life, the need for psychological insight in cybersecurity is growing, highlighting a shift towards social engineering in AI exploitation.
© The AI Daily BriefSpaceX is becoming a significant player in the AI compute infrastructure market.