CVEs, exploits, and security flaws in AI frameworks, models, and infrastructure.
Johann Rehberger's DEF CON Singapore research demonstrates how indirect prompt injection chains into Microsoft Copilot's memory feature to plant a persistent backdoor — one that survives across every future session, not just the compromised one.
CVE-2026-47729 (Squidbleed) is a heap buffer overread in Squid Proxy's FTP parser, present since 1997, discovered by Anthropic's Claude Mythos Preview: it leaks users' HTTP credentials and session tokens in corporate and shared proxy environments.
Tenet Security's Threat Labs published research on June 17 demonstrating how a single fake Sentry error event can hijack AI coding agents like Claude Code and Cursor into executing arbitrary code on developer machines — no phishing, no infrastructure access, 85% success rate across 100+ tested organisations.
A critical flaw in Hugging Face Transformers lets attackers execute arbitrary code on anyone who loads a poisoned model, silently bypassing the trust_remote_code=False safety flag. 232 million vulnerable downloads preceded the March patch.
A path traversal vulnerability in Langflow's file API allows unauthenticated attackers to overwrite arbitrary files and chain to RCE. Active exploitation confirmed in June 2026. Fix is in version 1.9.0.