SplxAI
Automated continuous red teaming and real-time AI firewall for agentic apps.
Acquired — still operating
now a Zscaler product line
About
SplxAI is an automated red-teaming and AI security platform for LLM and agentic applications, now a Zscaler product line. It simulates thousands of adversarial attack scenarios across text, image, voice, and documents, delivers dynamic risk analysis with remediation guidance, and doubles as a real-time AI firewall that monitors inputs and outputs to block threats with near-zero latency.
It maps findings to NIST AI RMF, OWASP LLM Top 10, EU AI Act, and ISO 42001, integrates with CI/CD, and partners with Databricks. Its founders came out of AI red-teaming at Cisco, Zscaler, and Wiz. Since the Zscaler acquisition it is best evaluated as part of an SSE platform stack rather than a standalone startup tool.
What it can do
- Continuous automated red teaming (multimodal)
- Real-time AI firewall with near-zero latency
- CI/CD pipeline integration
- Compliance mapping (NIST, OWASP, EU AI Act, ISO 42001)
- System prompt hardening and risk reports
Setup
- Contact Zscaler/SplxAI for platform access.
- Connect your LLM app or agent endpoints.
- Configure domain-specific test scenarios.
- Run continuous red teaming and review risk reports.
- Enable the real-time AI firewall for runtime protection.
Prerequisites
Enterprise account; access to the AI applications under test.
Use cases
Works with
Trust score — 6/10
Scored 6/10 (metrics tier, metrics-v1). Evidence: Now a Zscaler product line (since Oct 2025); evaluate as part of SSE stack; Founders from AI red-teaming at Cisco, Zscaler, Wiz; customers include KPMG, Glean; Multimodal coverage (text/image/voice) and CI/CD integration claimed; Commercial SaaS; attack-library depth is provider-stated. License: Proprietary SaaS.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 6.0/10
- Maintenance · recency, releases, issue load
- 7.0/10
- Openness · license + docs
- 4.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · Now a Zscaler product line (since Oct 2025); evaluate as part of SSE stack
- · Founders from AI red-teaming at Cisco, Zscaler, Wiz; customers include KPMG, Glean
- · Multimodal coverage (text/image/voice) and CI/CD integration claimed
- · Commercial SaaS; attack-library depth is provider-stated
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.