LatticeFlow AI
Evaluation platform from ETH Zurich researchers that maps frameworks to ready-made evaluations and adds automated agent red teaming
Quick Overview
| Company | LatticeFlow UK Ltd (LatticeFlow AI) |
|---|---|
| Category | AI Security Platforms |
| Headquarters | Zurich, Switzerland |
| Founded | Not stated |
| EU Presence | Yes - Switzerland and United Kingdom |
| Data Location | Not stated on the pages reviewed |
| Open Source | No |
| GDPR Compliant | Yes |
| Self-Hosting | Not stated |
| Main Features | AI Atlas frameworks, 100+ ready-to-run evaluations, YAML evaluation specs, synthetic datasets, agent red teaming, risk scoring, GRC sync, CLI and SDK |
| Pricing | Pricing on request |
| Best For | AI risk teams that need technical evidence tied to named frameworks |
| Replaces | HiddenLayer attack simulation, manual AI governance checklists |
Detailed Review
LatticeFlow treats security as one module of a wider AI risk platform. Atlas holds mapped frameworks, Evaluate runs the tests, Secure runs adversarial checks and Govern tracks risk and actions.
The privacy policy names LatticeFlow UK Ltd, and the footer lists a Zurich office. The About page says the founders are professors and researchers from ETH Zurich.
Frameworks as test suites
AI Atlas is described as a public registry of 40+ frameworks, 400+ controls and 100+ risks, each mapped to evaluations. Teams can start from OWASP, NIST RMF or ISO 42001 and edit them.
Security checks for agents
Secure runs an adaptive red teaming agent against data leakage, unauthorised actions, privilege abuse, goal hijacking and denial-of-wallet, aligned with OWASP and MITRE ATLAS, and returns mitigation suggestions.
Declarative evaluations
Evaluations are written as YAML so the method is explicit and auditable. Synthetic test data is generated from an application's knowledge base, documentation or code.
Limitations to Consider
This is governance and evaluation first, with security as one part, so it does not filter live prompts. The footer shows offices in Zurich, San Francisco, Dublin and Sofia, so check which entity signs your contract. No pricing is published. Overlaps with Giskard on agent testing.
Who Should Use LatticeFlow AI
Risk and compliance leads who must show regulators or auditors which framework control was tested and with what result.
Pros and Cons
Pros
- Frameworks mapped to runnable tests
- OWASP and MITRE ATLAS aligned red teaming
- YAML specs keep methods auditable
- GRC tool sync
- Academic roots at ETH Zurich
Cons
- Not an inline guardrail
- Pricing unpublished
- Contracting entity needs confirming
- Broader than a pure security tool
Alternatives to LatticeFlow 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 attack simulation on our alternatives page.
Frequently Asked Questions
Partly. Its Secure module runs red teaming and system-level checks, but the platform also covers evaluation and governance.
The site names OWASP, NIST RMF and ISO 42001 among 40+ frameworks in AI Atlas.
In declarative YAML specs, which the vendor says keeps the methodology explicit and auditable.