<h2>Must know</h2>
<ul>
<li><strong>AI agent security vulnerabilities are critical</strong>: Snyk’s ToxicSkills research found that 13.4% of AI agent skills have critical security issues, with 76 confirmed malicious payloads designed for credential theft and data exfiltration. This highlights a significant supply chain risk in AI agent development, affecting platforms like OpenClaw and Claude Code. Operant AI launched Agent Protector to offer real-time security for autonomous AI agents, emphasizing that security should be embedded by platform providers, not left to customers. <a href="https://app.daily.dev/feed-by-ids?id=0BdgmYNfg&amp;id=0e28QTlYb&amp;id=2vNCw5dB4&amp;id=dL9ys7HAk&amp;id=the2EenMO&amp;id=YLMZjOxxP&amp;id=Kcn4LiMDb&amp;id=Le8LdoPAn&amp;id=faybC0Zc5&amp;id=HQBjOB0ei" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Hardware shortages impact tech giants</strong>: Valve has delayed pricing and shipping for its Steam Machine and Steam Frame VR headset due to memory and storage shortages, driven by AI data center demand. RAM prices have more than doubled since October 2025. Sony’s CFO also addressed ongoing memory shortages and their potential impact on PS5 pricing, indicating a broader industry challenge affecting hardware availability and costs. <a href="https://app.daily.dev/feed-by-ids?id=iXYPYa4qJ&amp;id=FNWuAuO4A" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Substack data breach exposes user info</strong>: Substack disclosed a data breach from October 2025 that exposed user email addresses, phone numbers, and internal metadata, remaining undetected for four to five months. While passwords and financial information were not compromised, a threat actor claims nearly 700,000 records were affected, raising significant privacy concerns and prompting Substack to warn users about phishing attempts. <a href="https://app.daily.dev/posts/myrJIoXYs" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Anthropic and OpenAI clash over AI ads</strong>: Anthropic plans Super Bowl ads mocking OpenAI’s introduction of advertising in ChatGPT, positioning its Claude AI as ad-free, while OpenAI CEO Sam Altman called Anthropic’s ads dishonest. <a href="https://app.daily.dev/posts/TmM2mNSSP" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft appoints engineering quality czar</strong>: Microsoft CEO Satya Nadella appointed Charlie Bell as engineering quality czar to improve engineering quality across all products, addressing issues like Azure outages and Windows patch problems. <a href="https://app.daily.dev/posts/NGytU6zYE" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI bot traffic surges, impacting publishers</strong>: AI bots now account for one visit for every 31 human visits, primarily for retrieval augmented generation (RAG), causing clickthrough rates to publisher sites to plummet to 0.27% and threatening their financial viability. <a href="https://app.daily.dev/posts/5O2tzW5oh" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AWS launches new EC2 instances</strong>: AWS launched new EC2 High Memory U7i-6TB instances and C8id, M8id, and R8id instances with custom Intel Xeon processors, offering significant performance and memory bandwidth improvements for various workloads. <a href="https://app.daily.dev/posts/JPt4mjvME" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenAI launches enterprise AI agent platform</strong>: OpenAI launched ‘Frontier,’ an end-to-end platform for enterprises to build, deploy, and manage AI agents, allowing them to connect to external data and applications with controlled access. <a href="https://app.daily.dev/posts/TGZyay3jK" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Laravel releases AI SDK for application development</strong>: Laravel released an official AI SDK in beta, providing a unified API for integrating AI capabilities from multiple providers like Anthropic, Gemini, and OpenAI into Laravel applications, supporting agents, memory, and structured outputs. <a href="https://app.daily.dev/posts/ND9RAWZ9f" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>GNOME Shell and Mutter 50 beta released</strong>: GNOME Shell and Mutter 50 beta releases introduce stable Variable Refresh Rate (VRR) support, improved frame scheduling, and enhanced discrete GPU detection, paving the way for the stable GNOME 50 release in March. <a href="https://app.daily.dev/posts/buHOzLdcP" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>Elixir is the ‘best language for AI’</strong>: An article claims Elixir is the ‘best language for AI’ based on an unnamed ‘Tencent study’ that supposedly found a 97.5% LLM completion rate. Because a single, unverified study from an undisclosed source is all you need to declare a language superior for an entire field. Sure, Elixir has its merits, but let’s not pretend a secret Tencent study is the new industry benchmark. <a href="https://app.daily.dev/posts/1D7tGyRwk" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>UK deepfake detection plan criticized</strong>: The UK government announced plans to develop a framework for evaluating deepfake detection technologies, partnering with Microsoft and academics. Cybersecurity expert Dr. Ilia Kolochenko, however, states this initiative is unlikely to produce systemic improvements, arguing that the real challenge is enforcement, not detection. So, the government is busy reinventing the wheel while the actual problem—getting bad actors to stop—remains untouched. Classic. <a href="https://app.daily.dev/posts/EQCYBDiDB" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents are the new insider threat</strong>: A survey of IT executives suggests 1.5 million AI agents lack monitoring, leading to security incidents. Security experts are comparing this to the Titanic disaster, warning that unmonitored AI agents are an insider threat. Because nothing says ‘critical security vulnerability’ like an unmonitored script, which we’ve definitely never had before AI. It’s not a new problem, just a new buzzword to attach to it. <a href="https://app.daily.dev/posts/RMzBzQnM3" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
