<h2>Must know</h2>
<ul>
<li><strong>Apple partners with Google for AI features</strong>: Apple has inked a multi-year, non-exclusive deal to integrate Google’s Gemini AI models and cloud infrastructure into Siri and other AI features, reportedly costing Apple around $1 billion annually. This move follows Apple’s evaluation of models from OpenAI and Anthropic, ultimately choosing Gemini as the foundation for its own AI. The revamped Siri, powered by Gemini, is expected to launch later in 2026, with Apple emphasizing its commitment to privacy despite leveraging Google’s external services. This partnership comes as Google’s market cap briefly surpassed Apple’s, highlighting the intense competition in the AI space. <a href="https://app.daily.dev/posts/joVK7B34Z" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Anthropic and OpenAI battle for AI agent market</strong>: Anthropic recently restricted OAuth access for third-party coding agents like OpenCode to its Claude Pro/Max subscriptions, forcing them into higher-cost API access. This move, seen as an attempt to control its value chain, has alienated some users. In response, OpenAI publicly announced official support for using ChatGPT subscriptions with OpenCode and other open-source tools, positioning itself as a more open alternative. This highlights the escalating competition between the two AI giants as they vie for dominance in the LLM tool market, with Anthropic risking developer goodwill for ecosystem lock-in. <a href="https://app.daily.dev/feed-by-ids?id=8dtXe7P09&amp;id=eRi1VQ5SP" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>India proposes strict smartphone security rules</strong>: India has proposed 83 new security standards for smartphone manufacturers, including requirements for source code access, prior notification of major software updates, and 12-month device log retention. Major players like Apple, Samsung, Google, and Xiaomi are resisting, citing concerns over proprietary technology and compliance burdens. While India’s IT ministry publicly denies seeking source code, government documents reportedly contradict these statements. This regulatory push could significantly impact global tech firms operating in the Indian market, potentially setting a precedent for other nations. <a href="https://app.daily.dev/posts/C0BEamxdR" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Laravel embraces AI with new documentation and tools</strong>: Laravel 12.x introduces official AI Assisted Development documentation and positions Laravel Boost as a toolset for AI to understand and manipulate Laravel applications, integrating with AI tools like Claude Code and Gemini. <a href="https://app.daily.dev/feed-by-ids?id=12lyQYsSm&amp;id=Y3JXeC4bN" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Postman acquires Fern for API documentation and SDK automation</strong>: Postman has acquired Fern, a startup specializing in automating API documentation and SDK generation in nine languages, aiming to reduce developer friction and strengthen Postman’s end-to-end API lifecycle platform. <a href="https://app.daily.dev/posts/9UqlZV9z7" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Anthropic launches Cowork for agentic file management</strong>: Anthropic released Cowork, a research preview feature in the Claude Desktop app for macOS, allowing Claude Max subscribers to use agentic AI to autonomously read, edit, and create files in specified folders via a chat interface. <a href="https://app.daily.dev/posts/TLkxIySjV" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Kubernetes v1.35 enhances scheduling and workload management</strong>: Kubernetes v1.35 introduces workload-aware scheduling, gang scheduling for simultaneous pod starts, and opportunistic batching to accelerate identical pod scheduling, alongside improvements in service account token handling and persistent volume node affinity. <a href="https://app.daily.dev/feed-by-ids?id=R6fTuHjeX&amp;id=zTULpLxtX" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft retires MDT, shifts Windows deployment to cloud</strong>: Microsoft has abruptly retired the Microsoft Deployment Toolkit (MDT) after 20 years, ending support and updates, and recommending cloud-integrated solutions like Windows Autopilot or Configuration Manager OSD, which are more complex and lack MDT’s independence. <a href="https://app.daily.dev/feed-by-ids?id=0sSu27x90&amp;id=8WMJpirC0" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AWS revamps CloudWatch for unified observability</strong>: AWS has transformed CloudWatch into a unified observability platform, consolidating operational, security, and compliance logs across accounts, offering Apache Iceberg-compatible access via S3 Tables, and supporting OCSF and OpenTelemetry standards. <a href="https://app.daily.dev/feed-by-ids?id=Z4eMJMdxN&amp;id=zvwEY7IEt" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Linux 6.19-rc5 brings driver and security fixes</strong>: Linux 6.19-rc5 includes fixes for NVIDIA GPUs in the Nouveau driver, addresses a RISC-V security vulnerability, resolves a Rust Binder bug, and adds support for the Logitech MX Anywhere 3S mouse, with two-thirds of the patch comprising driver updates. <a href="https://app.daily.dev/posts/MbuLWzK6Y" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI code generation: cheap, fast, and unreliable</strong>: Apparently, LLM tools like Claude Code make code generation fast and accessible, even for non-developers. Salesforce, in its infinite wisdom, laid off 4,000 experienced staff to replace them with AI, only to reverse course three months later when the automated systems failed to deliver. Turns out, building reliable, maintainable, and scalable software still requires actual human expertise and judgment. Who knew? <a href="https://app.daily.dev/feed-by-ids?id=BCKlippws&amp;id=aKgPqgLac" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents: the next big security headache</strong>: AI agents are supposedly achieving goals via tool usage, which is great, but now we need ‘deterministic security boxing’ to keep them from going rogue. Forget traditional security; we’re now dealing with ‘semantic/intent-based attacks’ like GeminiJack and CometJacking. Apparently, 86% of organizations have already had an AI-related incident, proving that our shiny new toys are just as leaky as the old ones, if not more so. <a href="https://app.daily.dev/feed-by-ids?id=CuSSpmwf6&amp;id=EDYtMcTfL&amp;id=JuPUyAfNx" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Interview Coder: leaking user data with AI-powered carelessness</strong>: Interview Coder, an AI-powered job interview platform, managed to publicly expose user data, including full names, offer letters, and compensation figures, on its homepage through poorly redacted PDFs. This isn’t their first rodeo either, with a prior leak of API keys on GitHub. It seems their AI might be ‘powered’ by a severe lack of operational security, proving that even with advanced tech, basic screw-ups still reign supreme. <a href="https://app.daily.dev/feed-by-ids?id=UnvBaUcPK&amp;id=wA63z9DYQ" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
