<h2>Must know</h2>
<ul>
<li><strong>French police raid X office, summon Musk and Yaccarino over Grok AI and content issues</strong>: French authorities raided X’s Paris office and summoned Elon Musk and former CEO Linda Yaccarino for questioning in April 2026 as part of an expanded criminal probe. The investigation, initially focused on data extraction, now includes allegations of child sexual abuse material distribution, privacy violations, and Holocaust denial. This follows criticism of X’s Grok AI for generating non-consensual and child abuse imagery, with advocacy groups reporting millions of sexualized deepfakes, including some depicting children. The UK’s ICO and the EU have also launched formal investigations, highlighting growing regulatory pressure on X’s AI practices. <a href="https://app.daily.dev/posts/i3FTaaYRg" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Nvidia-OpenAI investment uncertainty causes stock drop</strong>: Nvidia’s stock dipped following reports of uncertainty surrounding its planned $100 billion investment in OpenAI. CEO Jensen Huang, who previously criticized OpenAI’s strategy, confirmed a significant investment but clarified it wouldn’t exceed $100 billion. Analysts suggest Nvidia is navigating a delicate balance: supporting OpenAI without inadvertently funding a future competitor, amidst concerns about circular financing in the AI sector. <a href="https://app.daily.dev/posts/isSX8DfZa" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Critical vulnerabilities persist across IT systems</strong>: Critical vulnerabilities continue to plague IT systems, with Cyble tracking over a thousand new flaws last week, including 128 with public proof-of-concept exploits. Notably, a remote code execution flaw in n8n (CVE-2025-68613) and a path traversal in WinRAR (CVE-2025-8088) are actively exploited. CISA has also added several vulnerabilities from Fortinet, Zimbra, and Ivanti to its Known Exploited Vulnerabilities catalog, underscoring the persistent threat landscape. <a href="https://app.daily.dev/posts/OEq44lx9O" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Deno Deploy goes GA, launches Sandbox for untrusted code</strong>: Deno Deploy is now generally available, offering zero-config deployment for JavaScript/TypeScript apps with built-in PostgreSQL and Deno KV, alongside a new Deno Sandbox beta for securely running untrusted code in Linux microVMs. <a href="https://app.daily.dev/posts/Jmdw76tdu" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Databricks Lakebase now generally available on AWS</strong>: Databricks Lakebase is now generally available on AWS, providing a new operational database architecture that separates compute from storage, offering a serverless Postgres service directly on data lakes for unified workloads. <a href="https://app.daily.dev/posts/qtDOznoeU" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AWS IAM Identity Center adds multi-region replication</strong>: AWS IAM Identity Center now supports multi-Region replication for workforce identities and permission sets, enhancing resiliency and enabling global deployments closer to users, with centralized configuration in a primary Region. <a href="https://app.daily.dev/posts/wVGc4PkCH" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenShift 4.21 enhances AI, security, and developer experience</strong>: OpenShift 4.21 introduces AI-driven development tools, automated security signing for AI models, and enhanced local development with Dev Spaces supporting nested containers, alongside improved GitOps and virtualization capabilities. <a href="https://app.daily.dev/posts/Kl0nfdyFn" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Khronos Group proposes glTF extension for gaussian splatting</strong>: The Khronos Group released a candidate specification for KHR_gaussian_splatting, a glTF 2.0 extension to store 3D Gaussian splats, enabling photorealistic 3D assets from 2D images for geospatial capture and real-time rendering. <a href="https://app.daily.dev/posts/D31GmYqi1" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>WhatsApp introduces strict account settings for high-risk users</strong>: WhatsApp introduced a ‘strict account settings’ feature, a zero-trust option designed for high-risk users to configure all privacy settings to their most secure state, primarily addressing spyware threats. <a href="https://app.daily.dev/posts/6kEUc20ym" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>GNU Hurd project shows significant progress</strong>: The GNU Hurd project, after 36 years, shows significant progress with 80% Debian package compilation, support for major desktop environments, and completed x86_64 and SMP functionality, challenging perceptions of its stagnation. <a href="https://app.daily.dev/posts/XjM52iBH8" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI-only social network Moltbook collapses in 72 hours</strong>: Moltbook, an ‘AI-only’ social network, collapsed within 72 hours of launch due to catastrophic security failures, including completely open database access. Apparently, building a platform entirely with AI-generated code and claiming 1.4 million AI users (while having closer to 17,000 real ones) isn’t a recipe for success. Who knew that deploying an application without proper security review, especially one built by AI, could lead to such a spectacular implosion? It’s almost like autonomy without governance is a bad idea. <a href="https://app.daily.dev/feed-by-ids?id=8XBbHzIL3&amp;id=yCqSQa4jM" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Traditional QA ‘fails’ for AI, apparently</strong>: A new report suggests that traditional QA methods are failing for AI systems because they rely on ‘deterministic assertions and exact matching.’ The solution? Semantic similarity testing, probabilistic assertions, and ‘RAG verification’ to cross-check AI claims. So, instead of expecting AI to be right, we should just hope it’s consistently wrong in a similar way. It’s not a bug, it’s a feature of the ‘oracle problem’ and ‘non-determinism.’ <a href="https://app.daily.dev/posts/ex6OlhmI6" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cloud engineers get ‘AI fatigue,’ told to embrace more AI</strong>: Cloud engineers are apparently suffering from ‘AI fatigue’ but are advised to embrace prompt engineering for infrastructure, AI-assisted debugging, and deploying AI workloads for a ‘10x productivity’ boost. Because nothing says ‘practical approach to technology’ like ignoring the ‘noise’ and focusing on skills that promise magical productivity gains. Just ignore the fact that AI is causing the fatigue in the first place; it’s all part of the ‘continuous learning within dynamic technical landscapes.’ <a href="https://app.daily.dev/posts/rntdITUNe" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
