RSAC 2026 Day 3–4 saw a flurry of vendor announcements, led by Accenture and Anthropic launching Cyber[.]AI for security operations, which uses Anthropic’s Claude model as a central reasoning engine and includes a library of autonomous agents plus an Agent Shield governance component.
Other highlights included Akamai and Bolster AI’s Brand Guardian to detect brand impersonation and phishing campaigns, and Arctic Wolf warning of an AI‑assisted malware surge, noting that 39% of samples initially evaded signature‑based detection.
Barracuda Networks unveiled AI‑driven risk visibility across its BarracudaONE platform and refreshed its Partner Success Program to support partners, while CrowdStrike introduced the Charlotte AI AgentWorks ecosystem for building secure agents with launch partners such as Accenture and OpenAI.
Additional items included DigiCert’s updates to Document Trust Manager for centralised key management and streamlined signing workflows, and Qualys releasing a remediation study on over 1 billion CISA KEV records showing vulnerability volumes rising 6.5x in three years and that 63% of critical vulnerabilities remained open on day seven in 2025.