PREVIEW. This is how it will look in an inbox; it was not sent to you. Each real email carries its own unsubscribe link and Keryn's postal address. KERYN HEADLINES Daily, Wednesday 8 April 2026 Retrospective Covers 7 April 2026 00:00 to 8 April 2026 00:00 UTC Web: https://keryn.dev/headlines/daily/2026-04-08/ TOP DEVELOPMENTS [1] Anthropic's Project Glasswing - restricting Claude Mythos to security researchers - sounds necessary to me Simon Willison · Research · https://simonwillison.net/2026/Apr/7/project-glasswing [2] ‘GrafanaGhost’ bypasses Grafana’s AI defenses without leaving a trace CyberScoop · Research · https://cyberscoop.com/grafanaghost-grafana-prompt-injection-vulnerability-data-exfiltration [3] SOHO router compromise leads to DNS hijacking and adversary-in-the-middle attacks Microsoft · Incident · https://microsoft.com/en-us/security/blog/2026/04/07/soho-router-compromise-leads-to-dns-hijacking-and-adversary-in-the-middle-attacks SAST [4] Guiding Symbolic Execution with Static Analysis and LLMs for Vulnerability Discovery arXiv · Research · https://arxiv.org/abs/2604.06506 [5] Prioritize security alerts with runtime context from Dynatrace GitHub · Analysis · https://github.blog/changelog/2026-04-07-prioritize-security-alerts-with-runtime-context-from-dynatrace SCA [6] Dependabot alerts are now assignable to AI agents for remediation GitHub · Analysis · https://github.blog/changelog/2026-04-07-dependabot-alerts-are-now-assignable-to-ai-agents-for-remediation [7] Attacks rewritten: Where malware enters the build Chainguard · Analysis · https://chainguard.dev/unchained/attacks-rewritten-where-malware-enters-the-build DAST [8] Why Your Automated Pentesting Tool Just Hit a Wall BleepingComputer · Analysis · https://bleepingcomputer.com/news/security/why-your-automated-pentesting-tool-just-hit-a-wall [9] Hackers or Hallucinators? A Comprehensive Analysis of LLM-Based Automated Penetration Testing arXiv · Research · https://arxiv.org/abs/2604.05719 AGENTIC SDLC [10] Broken by Default: A Formal Verification Study of Security Vulnerabilities in AI-Generated Code arXiv · Research · https://arxiv.org/abs/2604.05292 REMEDIATION [11] Does Pass Rate Tell the Whole Story? Evaluating Design Constraint Compliance in LLM-based Issue Resolution arXiv · Research · https://arxiv.org/abs/2604.05955 TRUST [12] Blockchain and AI: Securing Intelligent Networks for the Future arXiv · Research · https://arxiv.org/abs/2604.06323 [13] Governance and Regulation of Artificial Intelligence in Developing Countries: A Case Study of Nigeria arXiv · Research · https://arxiv.org/abs/2604.06018 [14] Auditable Agents arXiv · Research · https://arxiv.org/abs/2604.05485 MODELS [15] Towards Securing IIoT: An Innovative Privacy-Preserving Anomaly Detector Based on Federated Learning arXiv · Research · https://arxiv.org/abs/2604.06101 [16] Adversarial Robustness of Time-Series Classification for Crystal Collimator Alignment arXiv · Research · https://arxiv.org/abs/2604.06289 [17] Adjustable Text-Guided Backdoor Attacks with Natural-Word Triggers on Multimodal Pretrained Models arXiv · Research · https://arxiv.org/abs/2604.05809 AI SECURITY [18] The Defense Trilemma: Why Prompt Injection Defense Wrappers Fail? arXiv · Research · https://arxiv.org/abs/2604.06436 [19] A Formal Security Framework for MCP-Based AI Agents: Threat Taxonomy, Verification Models, and Defense Mechanisms arXiv · Research · https://arxiv.org/abs/2604.05969 [20] ClawLess: A Security Model of AI Agents arXiv · Research · https://arxiv.org/abs/2604.06284 ADVISORIES IN THIS WINDOW 65 npm advisories · 38 critical · 3 high · 22 moderate · 2 low 11 Maven advisories · 1 critical · 4 high · 5 moderate · 1 low 7 Go advisories · 2 critical · 4 high · 1 moderate 6 PyPI advisories · 1 critical · 2 high · 3 moderate 2 crates.io advisories · 2 high 1 Packagist advisory · 1 moderate All advisories: https://keryn.dev/headlines/advisories/ Titles, links and short excerpts only; every item links to its source. Preview: the unsubscribe link and Keryn's postal address are added to each email when it is sent.