Anthropic touts AI cybersecurity project with Big Tech partners

Anthropic partnered with Accenture to launch Cyber AI, a Claude-powered cybersecurity platform automating threat assessment, triage and remediation across enterprise networks. The solution integrates Anthropic's Claude models as reasoning engine with Accenture's agent library, compressing attack response timelines from weeks to hours. Initial deployments cut scan times from days to under 60 minutes while boosting coverage from 10% to 80%.

Accenture already runs Cyber AI internally across 1,600 applications and 500,000 APIs, eliminating critical vulnerability backlogs and improving service delivery 35%. A Fortune 500 agriculture client accelerated identity migrations through automated access management at scale. The platform includes Agent Shield for real-time monitoring of autonomous AI agents within defined risk parameters.

Launch follows Fortune's revelation of Anthropic's secret Claude Mythos model described internally as "far ahead of any other AI in cyber capabilities" which simulates attacks outpacing human defenders. Cybersecurity stocks plunged 5-11% on the disclosure, with CrowdStrike, Palo Alto and Zscaler hit hardest as markets priced AI-driven offense/defense asymmetry.

Partnership positions Anthropic against Palo Alto Networks' use of Claude 3.5 on AWS for code optimization and Infosys' enterprise deployments. Google, Amazon, Apple and Microsoft backed Anthropic's lawsuit challenging Pentagon blacklisting after the firm refused mass surveillance/autonomous weapons uses.

Anthropic CEO Dario Amodei frames cybersecurity as agentic AI's highest-ROI application, with Claude reasoning across vast datasets while respecting governance boundaries. Accenture reports adversaries leverage AI to compress attack timelines, demanding machine-speed defenses.

Timing coincides with Anthropic's $200 million private equity venture talks with Blackstone and Hellman & Friedman targeting portfolio company AI sales. Cyber AI enters $250 billion market as 9-in-10 organizations cite AI vulnerabilities as fastest-growing threat.