AI’s New Contest: Access, Agents and Custom Chips

Anthropic is opening Claude Mythos 5 to enterprise repository scanning while keeping the underlying cyber model behind a restricted interface—a useful test of how frontier access may be governed. Jesse also examines an exploited maximum-severity Entra ID flaw, Alibaba’s costly AI-cloud expansion, Waymo’s first purpose-built 5-nanometre driving chip, OpenAI’s national-security oversight initiative, urgent Cisco fixes, and Supermicro’s export-compliance investigation. In What Changes for You: Google brings Antigravity’s agentic development tools into eligible Gemini Enterprise plans, with central controls and substantial compliance limits. Full transcript: https://www.owenonthenet.com/podcast/2026-08-22/ Sources: https://claude.com/blog/bringing-claude-mythos-5-to-more-defenders https://news.bloomberglaw.com/us-law-week/anthropic-makes-claude-mythos-5-available-for-security-scans https://msrc.microsoft.com/update-guide/vulnerability/CVE-2026-69836 https://www.bleepingcomputer.com/news/microsoft/microsoft-warns-of-max-severity-entra-id-flaw-exploited-in-attacks/amp https://www.alibabagroup.com/en-US/document-2026456290057781248 https://apnews.com/article/8a30302d23a96fc7b9aab664b9c1897d https://waymo.com/blog/2026/08/look-under-our-trunk https://openai.com/index/strengthening-democratic-oversight-in-national-security https://sec.cloudapps.cisco.com/security/center/content/CiscoSecurityAdvisory/cisco-sa-hardening-csw1-shSvndWP https://ir.supermicro.com/news/news-details/2026/Supermicro-Announces-Completion-of-Independent-Investigation-and-Continued-Enhancement-of-Export-Compliance-Program/default.aspx https://assets.theregister.com/2026/08/21/20261?td=keepreading https://cloud.google.com/blog/products/ai-machine-learning/expanding-google-antigravity-for-enterprise-customers https://docs.cloud.google.com/gemini/enterprise/docs/ai-developer-tools-overview