{"uuid": "ce89a588-cb8f-4bb0-b254-2758cfd49a55", "vulnerability_lookup_origin": "1a89b78e-f703-45f3-bb86-59eb712668bd", "author": "9f56dd64-161d-43a6-b9c3-555944290a09", "vulnerability": "CVE-2026-100741", "type": "seen", "source": "https://gist.github.com/ichintu/295c47a4fdd1542fe1dda3872969c6bd", "content": "This daily threat intelligence update covers emerging AI security practices and critical software vulnerabilities:\n\n\u2022 **AI Guardrails:** Teams are developing adaptable, general-purpose proxy-layer frameworks to detect LLM threats like prompt injection and jailbreaks. Implementation faces a trade-off between evaluation metrics and documented capabilities, particularly regarding jailbreak detection and language support.\n\n\u2022 **wolfSSL (3 HIGH/8.3 CVEs):** Multiple flaws undermine TLS/DTLS certificate verification and authentication. Issues include ignoring public keys in trusted peer validation (CVE-2026-93302), accepting unsolicited RawPublicKey certificates (CVE-2026-89136), and incorrectly validating OCSP stapling issuers (CVE-2026-89102), all enabling authentication bypasses and certificate forgery.\n\n\u2022 **hMailServer (CRITICAL/9.8):** An Eval Injection vulnerability (CVE-2026-100741) allows unauthenticated remote code execution via JScript event scripting when exploited through specific SMTP/POP3/IMAP login sequences.\n\n\u2022 **Citrix NetScaler:** Two unpatched zero-day RCE flaws are actively exploited in the wild, prompting some administrators to preemptively take affected appliances offline.", "creation_timestamp": "2026-09-27T12:00:40.000000Z"}