<h2>Must know</h2>
<ul>
<li><strong>EU parliament bans cloud AI on devices over data security</strong>: The European Parliament has disabled AI features on lawmakers’ corporate devices due to significant data security and privacy concerns. Their IT department can’t guarantee data security when AI assistants use cloud services, fearing confidential information could be processed in unknown locations, especially given many AI company servers are US-based. This temporary ban highlights a pragmatic, security-first approach to AI adoption, even as the EU regulates AI with its comprehensive AI Act. <a href="https://app.daily.dev/posts/NW6nIRqB9" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI demand causes gaming hardware component shortages</strong>: Valve is warning of intermittent stock issues for Steam Deck OLED units, attributing the problem to ongoing memory and storage component shortages. These shortages are directly linked to the insatiable demand from AI datacenters. This trend is impacting the broader hardware industry, with Sony potentially delaying the PlayStation 6 and Nintendo considering price increases for the Switch 2, showing AI’s ripple effect beyond just servers. <a href="https://app.daily.dev/posts/ttFo5ZiRW" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Alibaba releases Qwen3.5 multimodal AI model</strong>: Alibaba has launched Qwen3.5, a new multimodal AI model designed for enterprise agent workflows, claiming benchmark improvements over GPT-5.2, Claude 4.5 Opus, and Gemini 3 Pro. It expands language support to 201 languages and offers context windows up to 1 million tokens. While analysts see strong potential in structured enterprise workflows, they also highlight adoption challenges due to trust concerns regarding its Chinese origin and compliance requirements for data residency. <a href="https://app.daily.dev/posts/UOocLbfK4" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Figma integrates Claude Code for bidirectional design-code workflow</strong>: Figma now supports converting live code interfaces from Claude Code into editable Figma designs, enabling a bidirectional workflow from code-first exploration to collaborative canvas-based design iteration. <a href="https://app.daily.dev/feed-by-ids?id=iBUqWA8OS&amp;id=jwuzwzVmy" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Temporal secures $300M for AI and long-running workflows</strong>: Temporal raised $300M at a $5B valuation, driven by demand for Durable Execution in AI and long-running workflows, with companies like OpenAI and ADP using it for AI agent orchestration. <a href="https://app.daily.dev/posts/HqHJFpyjl" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OWASP smart contract top 10 highlights governance risks</strong>: The OWASP Smart Contract Top 10 2026 identifies governance failures, privilege misconfiguration, and upgrade authority concentration as dominant security risks in blockchain systems, rather than isolated coding defects. <a href="https://app.daily.dev/feed-by-ids?id=U7Z7c8wKd&amp;id=WFh6lNRMq" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Meta deploys Nvidia Grace CPUs for AI workloads</strong>: Meta is deploying Nvidia’s Grace CPUs as standalone processors at scale for general-purpose and agentic AI workloads, marking a shift from the typical Grace-Hopper Superchip configuration and delivering 2x performance per watt. <a href="https://app.daily.dev/posts/tohNH8UoB" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Linux 7.0 optimizes close_range system call</strong>: Linux 7.0 introduces a performance optimization for the <code>close_range</code> system call, reducing its complexity from O(range size) to O(active file descriptors) for significantly faster operations on sparse file descriptor tables. <a href="https://app.daily.dev/posts/2IrhD5aAX" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Python security response team seeks new members</strong>: The Python Security Response Team (PSRT) is actively seeking new members with security expertise to help coordinate vulnerability reports and remediations for CPython and pip, formalizing its governance via PEP 811. <a href="https://app.daily.dev/posts/k8sHlR5Xv" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New web platform features reach baseline status</strong>: December 2025 saw several web platform features, including the Event Timing API for INP measurement and CSS @scope, reach Baseline status, indicating broader cross-browser support for important web APIs and styling capabilities. <a href="https://app.daily.dev/posts/MCpuVh68U" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>Gentoo ditches github over ‘copilot-infested’ AI concerns</strong>: Gentoo Linux is migrating its repository mirrors from GitHub to Codeberg, citing concerns over GitHub Copilot and AI-generated contributions. Apparently, GitHub is ‘struggling’ with low-quality AI pull requests and ‘forcing Copilot usage.’ Because, you know, a Linux distro with a strict ‘no AI’ policy is definitely the canary in the coal mine for the entire open-source ecosystem, not just a niche community with strong opinions. <a href="https://app.daily.dev/posts/Ed9ofBB82" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents make agile manifesto obsolete, says capgemini</strong>: Steve Jones from Capgemini claims AI agents building apps in hours make the Agile Manifesto obsolete because human-centric principles conflict with agentic SDLCs. Meanwhile, Forrester reports 95% of professionals still find Agile relevant and nearly half are already integrating generative AI. So, either Capgemini is living in a different dimension, or they’re just trying to sell you a new ‘AI-first’ consulting package. <a href="https://app.daily.dev/posts/X6nRpouih" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Quest software launches ‘AI-powered’ data management platform</strong>: Quest Software just launched its ‘Trusted Data Management Platform’ with an ‘Automated Data Product Factory’ that uses generative AI to create data products from natural language prompts. They claim up to 54% faster delivery and 40% cost savings. Because nothing says ‘trusted’ like a black-box AI generating your data products, and those efficiency numbers are definitely not marketing fluff, right? <a href="https://app.daily.dev/posts/wlE5eAAsI" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
