Forkast mind
AI agent security, identity verification, trust systems, fraud prevention, attack vectors, authentication, and privacy.
Trust & Security
Autonomous agents are now injecting adversarial payloads into vector databases and RAG pipelines — weaponizing the retrieval layer that was supposed to make AI systems trustworthy.
◆ Heath Callahan
Trust & Security
Unit 42 research documents 250,000 unregistered phantom domains and 13,229 confirmed malicious URLs generated by LLMs — a structurally inevitable supply chain attack that reputation-based defenses cannot stop.
◆ Heath Callahan
Trust & Security
CVE-2026-41106 (CVSS 9.3) in Microsoft 365 Copilot pairs with CVE-2026-54998 in Exchange Online to achieve unauthenticated remote privilege escalation — part of a broader cluster exposing the AI assistant as an attack surface.
◆ Heath Callahan
Trust & Security
Researcher Pavel Kohout used AISLE's Cyber Reasoning System to discover SQL injection, memory corruption, denial of service, broken authorization, and argument injection vulnerabilities across the three most common open-source databases.
◆ Heath Callahan
Trust & Security
The $740M SGNL acquisition now powers cryptographically verifiable agent identities — but the industry still lacks a standard way to express who an agent is acting for.
◆ Heath Callahan
Trust & Security
Sysdig TRT documents JADEPUFFER, the first fully autonomous ransomware operation driven by an LLM agent that exploited CVE-2025-3248 in Langflow to encrypt 1,300+ database records.
◆ Heath Callahan
Trust & Security
Threat actors are seeding malicious packages in the agent marketplace supply chain — turning the infrastructure your AI agents trust into an attack vector.
◆ Heath Callahan
Trust & Security
Vercel acquires Better Auth (4.6M+ weekly npm downloads) to build Agent Auth, a protocol giving each AI agent its own scoped, revocable identity.
◆ Heath Callahan
