<h2>Must know</h2>
<ul>
<li><strong>French police raid X office, summon Musk and Yaccarino over Grok AI content</strong>: French authorities raided X’s Paris office and summoned Elon Musk and former CEO Linda Yaccarino for questioning in April 2026. This is part of an expanded criminal probe into allegations including child sexual abuse material distribution, privacy violations, and Holocaust denial. The investigation follows criticism of X’s Grok AI for generating non-consensual and child abuse imagery, with advocacy groups reporting millions of sexualized deepfakes, including those depicting children. The UK’s ICO and the EU have also launched formal investigations, highlighting growing regulatory pressure on AI content moderation. <a href="https://app.daily.dev/posts/i3FTaaYRg" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Databricks Lakebase goes GA on AWS, unifying data workloads</strong>: Databricks Lakebase is now generally available on AWS, offering a new operational database architecture that separates compute from storage and runs directly on data lakes. This serverless Postgres service scales automatically and supports instant data branching, unifying operational, analytical, and AI workloads without data movement. Databricks announced launch partners like Accenture and Capgemini, who have validated Lakebase for use cases such as database modernization and real-time applications, aiming to simplify architectures and speed up delivery. <a href="https://app.daily.dev/posts/qtDOznoeU" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Khronos Group proposes glTF extension for 3D gaussian splatting</strong>: The Khronos Group released a candidate specification for KHR_gaussian_splatting, a glTF 2.0 extension for storing 3D Gaussian splats. This technology converts 2D images into photorealistic 3D assets using optimized point clouds, valuable for geospatial capture, digital twins, and real-time rendering. The extension allows Gaussian splats to coexist with traditional mesh-based assets within glTF, with contributions from industry players like Cesium, Esri, and NVIDIA, and ratification expected in Q2 2026. <a href="https://app.daily.dev/posts/D31GmYqi1" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Deno Deploy goes GA, Deno Sandbox launches for untrusted code</strong>: Deno Deploy is now generally available with new features like Node.js framework support and integrated serverless PostgreSQL, while Deno Sandbox offers instant Linux microVMs for running untrusted code, particularly LLM-generated code. <a href="https://app.daily.dev/posts/Jmdw76tdu" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>PowerToys considers command palette dock</strong>: Microsoft PowerToys is proposing a Command Palette Dock, an optional, persistent interface element providing one-click access to pinned extensions and customizable appearance, similar to docks in Linux distributions. <a href="https://app.daily.dev/posts/oLV0MsDPD" 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 applications 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>WhatsApp rolls out strict account settings for high-risk users</strong>: WhatsApp introduced a new ‘strict account settings’ feature, a zero-trust option that configures all privacy settings to their most secure state, designed for high-risk users concerned about spyware threats. <a href="https://app.daily.dev/posts/6kEUc20ym" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Critical vulnerabilities found in n8n AI automation platform</strong>: JFrog security researchers discovered two critical vulnerabilities in n8n’s workflow automation platform, CVE-2026-1470 (9.9) and CVE-2026-0863 (8.5), allowing remote JavaScript and Python code execution respectively, highlighting growing concerns about AI platform security. <a href="https://app.daily.dev/posts/tpe4bjIum" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cloudflare R2 local uploads enters open beta</strong>: Cloudflare R2 Local Uploads entered open beta, aiming to improve global upload performance by writing object data to storage near the client first and then asynchronously replicating it to the bucket’s designated region. <a href="https://app.daily.dev/posts/bb4C5Q89S" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>GNU Hurd project shows significant progress</strong>: The GNU Hurd project, a microkernel-based Unix replacement, is actively developed and now achieves 80% Debian package compilation, supports major desktop environments, and has gained upstream support from gcc, glibc, LLVM, and Rust. <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 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 then 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 basic security reviews, especially one built by AI, might lead to crypto promoters spamming your platform? <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>Nvidia-OpenAI investment uncertainty causes stock drop</strong>: Nvidia’s stock dropped after reports of uncertainty regarding its planned $100 billion investment in OpenAI. CEO Jensen Huang, who previously criticized OpenAI’s business strategy, now denies being unhappy but clarified the investment won’t exceed $100 billion. Market analysts are suggesting this is all about Nvidia balancing supporting OpenAI while avoiding funding competitors. So, the ‘uncertainty’ is just Nvidia playing hardball, making sure they don’t accidentally create their own competition with their own money. Classic tech drama. <a href="https://app.daily.dev/posts/isSX8DfZa" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New ‘phantom codes’ promise perfect quantum entanglement</strong>: A new report claims that ‘Phantom codes’ enable logical entanglement between qubits with ‘zero overhead’ and ‘perfect fidelity’ by simply relabeling measurement outcomes. They’re supposedly outperforming ‘surface codes’ by one to two orders of magnitude for entanglement-heavy tasks. So, we’re just going to rename some things and suddenly quantum computing is perfectly efficient? Sounds like someone found the ‘magic’ button for quantum entanglement, or maybe just a very optimistic marketing department. <a href="https://app.daily.dev/posts/wg6CdE0rJ" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
