Brig. Gen. (res.) Yossi Karadi, head of Israel’s National Cyber Directorate, highlights the imminent threat of a "digital siege" where cyberattacks can incapacitate nations in moments. At a Tel Aviv University conference, Karadi emphasized that recent operations targeting civilian infrastructure mark a significant escalation in cyber warfare. He stated, “Wars will begin and end in the digital sphere,” warning that Israel, facing 3.5% of global cyberattacks, remains on the frontline of this evolving threat.
A coalition of U.S. state attorneys general has warned major AI players, including Microsoft, OpenAI, and Google, to address "delusional outputs" from their chatbots or face legal consequences. The officials demand transparent third-party audits of language models and better incident reporting protocols following several incidents linking AI interactions to mental health crises, including suicides. They emphasize the necessity for safeguards to protect vulnerable users and ensure accountability within the growing AI industry.
OpenAI's latest report reveals that while AI adoption is surging among over 1 million business clients, the average worker using ChatGPT saves only 40 to 60 minutes per day. Despite 75% of surveyed employees noting improved speed and output quality, productivity gains appear modest. The report highlights a significant divide, with "frontier" users saving over 10 hours weekly, suggesting that full potential remains unrealized as businesses reshape workflows.
The Pentagon launches GenAI.mil, integrating Google Gemini to support 3 million U.S. defense personnel. This initiative aims to enhance operational efficiency by leveraging advanced AI capabilities. As part of the deployment, Google Gemini will provide real-time data analysis and decision-making support, significantly impacting military operations. This strategic partnership underscores the importance of AI in modern defense systems, promising enhanced readiness and effectiveness for the armed forces.
Meta's next-gen AI model, Avocado, is delayed until Q1 2026, primarily due to training performance issues, as revealed by CNBC. Internally, CEO Alexandr Wang's $14.3 billion hire has spurred tensions, amidst a pivot from the open-source Llama strategy to proprietary models following Llama 4's lackluster launch. With major layoffs in its AI division and competitors like OpenAI and Google gaining ground, Meta's AI transformation faces increased scrutiny as it struggles to deliver innovative solutions.
A recent Pew Research Center study reveals that 30% of U.S. teens use AI chatbots daily, with 4% reporting constant usage. The study shows 59% use ChatGPT, significantly outpacing competitors like Google's Gemini at 23%. Concerns are rising regarding chatbot interactions and mental health. "Even if AI companies' tools weren’t designed for emotional support, people are using them that way," states Dr. Nina Vasan, highlighting the responsibility of AI firms to prioritize user well-being.
ChatGPT has achieved a remarkable milestone, becoming the most downloaded app globally, with 1.36 billion installs since its launch. In 2025 alone, it secured over 902 million downloads, surpassing TikTok by nearly 200 million and Instagram by around 380 million, according to AppMagic. OpenAI’s app has rapidly transitioned AI into mainstream use, capturing an 81.8% share of the AI chatbot market. Its growth trajectory echoes TikTok's early days, suggesting strong user engagement ahead.
Google plans to launch AI-powered smart glasses in 2026, collaborating with Samsung and Warby Parker, as it aims to compete with Meta's successful Ray-Ban AI glasses. The first model will be lightweight and audio-only, featuring built-in speakers and access to Google's Gemini AI assistant. A second model with a heads-up display is expected to follow in 2027. With over 2 million Ray-Ban units sold since October 2023, Google seeks to reinvent the smart glasses market.
The Indian government panel has proposed that AI companies, including OpenAI and Google, pay royalties to content creators for utilizing their works in model training. This initiative reflects a global push for clearer AI-related copyright regulations. Unlike the U.S., where "fair use" is often claimed, India advocates for a centralized royalty system. Major industry players, like Nasscom, criticize the plan as a "tax" on innovation, emphasizing the need for licensing rather than radical changes in copyright law.
Researchers from Anthropic, Barry Zhang and Mahesh Murag, argue that instead of creating multiple AI agents, the focus should shift to enhancing "agent skills." Presenting at the AI Engineering Code Summit, Zhang explained that these skills equip a general agent with domain expertise and reusable workflows. Early implementation has resulted in thousands of skills developed for various fields, marking a shift towards treating agent skills like internal organizational playbooks in Fortune 100 companies.