Nearly two-thirds of top private AI companies have leaked sensitive information including API keys and credentials on GitHub. The affected firms represent over $400 billion in collective valuation, revealing serious security gaps in the AI boom.
Major tech companies lost $1.2 trillion in market value last Friday as analysts grow increasingly concerned about AI valuations. OpenAI executives made and then retracted comments about government loan guarantees for AI infrastructure. The industry appears to be entering a moment of truth.
DeepSeek’s senior researcher Chen Deli warned that AI will eventually replace most human work, outlining a three-stage framework for how companies should respond. His comments come amid growing global concern about AI’s trajectory and social impact.
DBS has upgraded its Joy chatbot with generative AI capabilities, and the results are impressive. The virtual assistant handled over 120,000 unique chats since February while boosting customer satisfaction scores by 23%. The bank is now planning to expand the service beyond Singapore.
Schools are deploying AI monitoring software to track student conversations with chatbots like ChatGPT and Character.ai. These systems flag concerning content but raise major privacy and effectiveness questions. The surveillance industry has quietly become a juggernaut in education technology.
The European Space Agency is partnering with Irish quantum startup Equal1 to install its Bell-1 quantum computer for Earth observation research. The six-qubit system represents ESA’s first quantum computing deployment and aims to accelerate processing of complex environmental data.
AI browsers are moving beyond simple search to actually executing tasks autonomously. With agentic AI capabilities, they can manage emails, plan trips, and handle repetitive work. The technology is shifting from business use to everyday consumer applications.
The LA Clippers’ new Intuit Dome features an AI experience called Connectopia that generates unique 8K worlds from text prompts in real time. Built with RTX 4090/5090 GPUs and running Gemma and LTX models, it represents one of the most advanced public AI installations anywhere. Every two weeks, the
A new side-channel attack can accurately guess what you’re discussing with AI chatbots like ChatGPT, even through encryption. The vulnerability exploits how chatbots stream responses word by word. Major providers are already implementing fixes.
XPeng’s CEO literally unzipped the robot to prove it wasn’t a human in a suit. The “Iron” unit features 22-degree flexibility hands and what the company claims is the industry’s first all-solid-state battery for home safety.
