AI breakthroughs, security risks, Anthropic pricing...

DeepSeek-R1, a prominent AI platform, has drawn scrutiny for potential security vulnerabilities. Experts have identified that its code output may become compromised when handling political topics, raising alarms about hidden censorship mechanisms and risks in enterprise applications. This brings to light the ethical and operational challenges associated with deploying AI systems in sensitive domains. Tech Radar

OpenAI's ChatGPT has merged its voice and text interfaces, allowing users to switch seamlessly between the two on a single screen. This update enhances conversational AI by making interactions more dynamic and responsive, further solidifying its position in the AI assistant market. TechCrunch

Cybersecurity firm Rubrik is testing an "undo button" for malfunctioning AI agents, addressing growing concerns about rogue AI behavior. This feature aims to restore data and mitigate risks posed to critical business operations, reflecting the increasing demand for safeguards in AI deployments. The Information

Chinese tech giant Alibaba announced that its Qwen AI chatbot achieved 10 million downloads in its first week of public beta, surpassing initial adoption rates of Western competitors like ChatGPT. The chatbot’s success highlights the strategic advantage Chinese companies hold due to exclusive access to a vast domestic market. Zdnet

Google is reportedly developing Aluminium OS, an Android-based operating system for laptops, which will feature deep AI integration via its Gemini model. This move signals a potential shift away from ChromeOS and underscores Google’s focus on AI-centric innovation. TechSpot

Anthropic has slashed prices for its Claude Opus 4.5 model by up to 67%, aiming to make its advanced AI tools more accessible to enterprises. Analysts note that this reflects an industry-wide push to democratize frontier AI capabilities while maintaining competitive differentiation. Computer World

Amazon has committed $50 billion to expand federal AI infrastructure in the U.S., focusing on high-performance computing (HPC) hardware and classified cloud regions. This ambitious investment underscores the pivotal role of AI in national security and scientific research. Tech Radar

Qualcomm has faced backlash from the tech community after rewriting Arduino's terms of service following its acquisition. Critics argue that the updated policies undermine open-source principles, sparking concerns about the future direction of the platform. TechSpot

President Trump has launched the Genesis Mission, a federal initiative to integrate AI with large datasets for scientific discovery. This program aims to enhance the U.S.’s global competitiveness in technology while fostering collaboration between national labs and AI developers. Scientific American The Register

A new study has explored how fragments of bacteria in the brain could influence sleep patterns. Researchers suggest that these microbial components may interact with neural mechanisms regulating rest, opening a novel avenue for understanding sleep disorders. Science Alert