Founded 2021

Adversa AI

Tel Aviv, Israel

Adversa AI runs an AI red-teaming platform that continuously tests AI agents, LLMs, and generative AI applications for security weaknesses. Its offering spans LLM threat modeling for risk profiling, an LLM vulnerability audit that runs continuous checks against known issues such as the OWASP LLM Top 10, and LLM red teaming that uses AI-enhanced attack simulation across a large library of mapped attack techniques to surface prompt injection, jailbreaking, and data-leakage risks.

Headquarters
Tel Aviv, Israel
Founded
2021
Website
adversa.ai/

Notable facts

  • Holds issued US Patent 11,275,841 ('A method and system of protecting an artificial intelligence (AI) application'), issued March 15, 2022, which the company markets as the world's first AI security patent.
  • Named an IDC Innovator in AI Security, 2024.
  • Published the AIRQ (AI Risk Quadrant) framework in 2026, an open-source AI agent security rating built with contributors from OWASP, CoSAI, CSA, and NIST, scoring 100+ agents.
  • Raised a pre-seed round in December 2021 from investors including Aviram Jenik, Moxxie Ventures, and VentureIsrael.

Service categories

Pricing

Pricing not publicly disclosed on the vendor's site. Typical for B2B AI-governance engagements, which are usually quoted per scope.

Case study

No self-published case study found for this vendor at time of research. This reflects what could be found publicly, not a claim about the vendor's actual client work.

Sources. This profile is compiled from public reference data, WebSearch-verified againstsource 1 · source 2 · source 3. Data as of 2026-07-13. See methodology.

Sourcing a consultant in this category today? We connect risk, compliance, and platform teams with qualified AI governance and compliance consultants.

Request consultants

Related vendors