Signal criticality: High
What happened: The Hacker News published "Critical Bifrost AI Gateway Flaw Lets Attackers Run Commands Without Credentials". A critical vulnerability in Bifrost, an open-source AI gateway that routes requests to more than 20 LLM providers, allows an unauthenticated attacker to run arbitrary commands on the gateway server with a single HTTP request. The flaw, tracked as CVE-2026-90898 (CVSS score: 9.8), affects all versions of the Bifrost HTTP transport before 2.1.0 when management authentication is The report describes a concrete compromise, exposure, or abuse pattern with direct defensive implications.
Key takeaways:
Original source: https://thehackernews.com/2026/09/critical-bifrost-ai-gateway-flaw-lets.html
Signal criticality: High
What happened: Dark Reading published "Prompt-Injection Bug Hits $4B Agentic AI App 'Manus'". AI apps that interpret external data (read: most AI apps) need exceptionally rigorous security filters, or attackers can take advantage The article focuses on governance, identity, guardrails, or permission boundaries around AI agents that can act with real system access. The practical question is what permissions, connected data, or follow-on actions this signal can influence in a real deployed workflow.
Key takeaways:
Original source: https://www.darkreading.com/application-security/prompt-injection-bug-agentic-ai-app-manus
Signal criticality: High
What happened: Help Net Security reported that latticeFlow AI offers managed risk assessments for enterprise AI systems LatticeFlow AI has announced the LatticeFlow AI Risk Center, an AI governance managed service that continuously assesses and controls AI risk, giving enterprises the technology, evidence, and expertise to scale AI with confidence and accelerate time to value. Enterprises decide which AI systems are assessed, the frameworks and risk thresholds that apply, the assessment frequency, and how identified risks are mitigated.
Key takeaways:
Original source: https://www.helpnetsecurity.com/2026/09/24/latticeflow-ai-risk-center/
Signal criticality: High
What happened: SecurityWeek reported that artificial Intelligence Kontext Security Emerges With $4 Million for AI Agent Runtime Controls The startup’s runtime enforcement platform evaluates AI agents in real time to provide visibility and control over their actions. By Ionut Arghire | September 24, 2026 (11:52 AM ET) Flipboard Reddit Whatsapp Whatsapp Email Kontext Security today launched publicly with $4 million raised in a funding round led by 42CAP, with additional support from a16z CSX and HTGF.
Key takeaways:
Original source: https://www.securityweek.com/kontext-security-emerges-with-4-million-for-ai-agent-runtime-controls/
Signal criticality: High
What happened: Microsoft Security Blog published "Reimagining the SOC for the agentic era in Microsoft Defender". We are announcing ISOC in Microsoft Defender: a foundation built for agentic security that brings leading solutions for SIEM and threat protection together The article focuses on governance, identity, guardrails, or permission boundaries around AI agents that can act with real system access. The practical question is what permissions, connected data, or follow-on actions this signal can influence in a real deployed workflow.
Key takeaways:
Original source: https://www.microsoft.com/en-us/security/blog/2026/09/23/reimagining-the-soc-for-the-agentic-era-in-microsoft-defender/
The strongest signal today is that AI security is being decided in the surrounding control layer — permissions, connectors, deterministic workflow design, response speed, and the infrastructure that still underpins trust. That is a more durable framing than generic agent hype, and it is the one worth carrying forward.