Adversa AI
Tel Aviv AI red teaming and coding agent security vendor that also publishes an open agent rating methodology
Quick Overview
| Company | Adversa AI Ltd |
|---|---|
| Category | AI Security Platforms |
| Headquarters | Tel Aviv, Israel |
| Founded | Not stated |
| EU Presence | Yes - Israel (counts as European for EP) |
| Data Location | Not stated on the pages reviewed |
| Open Source | No |
| GDPR Compliant | Yes |
| Self-Hosting | Not stated |
| Main Features | AI red teaming for LLM, agentic, generative AI and MCP, coding agent security, AI detection and response, assessment and assurance services, EU AI Act, DSA and DMA compliance pages |
| Pricing | Pricing on request (demo) |
| Best For | Security teams governing coding agents and teams buying external red teaming |
| Replaces | HiddenLayer AI red teaming |
Detailed Review
Adversa AI splits its offer between red teaming and a newer coding agent security platform. The red team pages cover LLMs, agentic AI, generative AI and MCP.
The privacy policy names Adversa AI Ltd with an address on Rothschild Boulevard, Tel Aviv, and the footer shows a 2026 copyright to Adversa AI Ltd. EP treats Israel as European since 22 October 2026.
Chains, not single alerts
The coding agent platform records what a developer agent reads and then what it runs. It judges the whole chain, so a single odd step is context rather than an alert, and it blocks dangerous chains before they finish.
Capture happens at the LLM traffic, the agent harness and the machine, which lets it work across several agent brands.
An open rating method
On 4 June 2026 the company published AIRQ, an AI Risk Quadrant rating of 100 enterprise AI agents with an open source methodology that buyers can reuse in procurement.
Compliance angle
The site has solution pages for the EU AI Act, the Digital Services Act and the Digital Markets Act, and it publishes a data processing agreement and a security and trust centre.
Limitations to Consider
The current homepage leads with coding agents, so a team seeking a general LLM application firewall will find less detail. Prices are unpublished and hosting is not described. Funding and awards on the site are vendor statements.
Who Should Use Adversa AI
CISOs and AppSec teams that need to control Claude Code, Copilot, Cursor and similar agents, or that want a specialist red team for LLM apps.
Pros and Cons
Pros
- Cross-vendor coding agent coverage
- Open AIRQ rating methodology
- Red teaming for MCP and agents
- DPA and trust centre published
- Compliance pages for EU rules
Cons
- Homepage focuses on coding agents
- No public pricing
- Hosting not described
- Needs endpoint sensor rollout
Alternatives to Adversa AI
Other European ai security platforms worth comparing before you decide:
Mindgard
United Kingdom · London AI red teaming that attacks, fixes and retests
NeuralTrust
Spain · Barcelona agent gateway and runtime guard with free tier
Advai
United Kingdom · UK AI assurance: choose, test and monitor AI systems
Giskard
France · Paris open-core red teaming and testing for AI agents
See all European ai security platforms or compare them directly against HiddenLayer AI red teaming on our alternatives page.
Frequently Asked Questions
AIRQ is an AI Risk Quadrant that rates 100 enterprise AI agents, published on 4 June 2026 with an open source methodology.
Yes, the site lists AI red teaming for MCP among its use cases.
The privacy policy gives a Tel Aviv address for Adversa AI Ltd. EP counts Israel as European.