Cisco AI Defense (Robust Intelligence)
AI security product line built on Robust Intelligence's algorithmic red-teaming and AI Firewall technology, now integrated into Cisco's security portfolio.
Visit Website ↗ + Add to Compare Claim This CompanyOverview
Robust Intelligence pioneered algorithmic red-teaming (automated, systematic adversarial testing of ML models) and built what it called the industry’s first AI Firewall, an inline control for detecting and blocking model-level attacks at runtime. Co-founded in 2021 by Yaron Singer, a Harvard computer science professor, and Kojin Oshiba, the company built a research-heavy reputation in the AI security field before being acquired.
Cisco acquired Robust Intelligence in October 2024 for a reported $400M, after the startup had raised about $44M and reached a valuation over $200M independently. The technology and team are now the foundation of Cisco AI Defense, an end-to-end AI security product announced roughly three months after the acquisition closed, which Cisco is positioning within its broader Security Cloud platform to protect AI application development and runtime use across enterprise customers.
Innovation Matrix Assessment
Went from a research-driven startup with a novel algorithmic red-teaming methodology to a fully productized, Cisco-branded platform (AI Defense) within about three months of the acquisition closing.
AI Firewall and red-teaming tooling give security teams concrete, inline controls, now distributed through Cisco's existing enterprise security-sales channel.
A $400M acquisition by a major networking/security incumbent, following independent VC validation at a $200M+ valuation, is a strong real-world signal.
Algorithmic red-teaming and an AI-specific inline firewall were genuinely new control categories when introduced in 2021, ahead of most of the market. However, Robust Intelligence is now Cisco AI Defense, part of Cisco (NASDAQ: CSCO), a public company with a market cap in the hundreds of billions (per nasdaq.com), which is the scale of an established incumbent rather than an emerging category disruptor, so this dimension is scored more conservatively.
Founder Yaron Singer's academic ML background and published red-teaming research give more independent grounding than most category peers, though Cisco has not published third-party efficacy validation of AI Defense specifically.
Now backed by Cisco's distribution and integrated into Cisco Security Cloud and Cisco Foundation AI, giving it durability tied to a large incumbent's roadmap.
Why CISOs Should Care
Provides an inline AI firewall and automated red-teaming that plug into an organization's existing Cisco security stack, lowering integration friction for security teams already standardized on Cisco.
What Makes It Different
Applies algorithmic (automated, systematic) red-teaming rather than manual pen-testing to continuously probe models for vulnerabilities, an approach that predates most competitors' red-teaming offerings.
The Matrix Verdict
70/100 — INCUMBENT
Strong Performer: genuine research pedigree and a large, credible acquirer, but as with other acquired startups in this category, independent post-acquisition efficacy data specific to Cisco AI Defense is not yet publicly available.
Editorial Note: Claims vs. Verified Findings
Acquisition price ($400M) and pre-acquisition funding (~$44M) are reported by trade press and Cisco's own blog posts, not a line-itemized SEC filing. Employee range reflects Cisco's total corporate headcount, not the AI Defense unit specifically, since Cisco does not break out unit-level staffing.
Sources
Alternatives to Cisco AI Defense (Robust Intelligence)
Adaptive Security
AI-driven platform that simulates deepfake, voice, and multichannel social-engineering attacks to train and test organizations against next-generation phishing.
Quilr
Early-stage agentic AI security startup building a 'Service-as-Software' platform to guard against human-related breaches and secure AI agent…
Zenity
Governance and security platform for AI agents and low-code/no-code development, securing agent identity, permissions and behavior across the…
Gray Swan AI
Carnegie Mellon-linked startup building automated red-teaming and runtime guardrails that frontier AI labs use to stress-test their models…
Tenzai
An agentic AI penetration testing startup building autonomous 'AI hackers' to find and validate exploitable vulnerabilities at a…
Alice (formerly ActiveFence)
Israeli AI security company (rebranded from ActiveFence in January 2026) offering a lifecycle platform to test, guard, and…