EVE AI Core
EVE AI Core
Lakera Guard is an ML-based AI security layer — detecting prompt injection, jailbreaks, and PII with proprietary trained detectors. EVE CoreGuard is a deterministic governance engine that enforces a regulatory policy and signs the decision. They solve adjacent but different problems: threat detection versus provable compliance enforcement.
Category: AI security / LLM guardrails (now part of Check Point).
Lakera is a well-regarded AI security vendor (a Gartner AI TRiSM Representative Vendor) whose Lakera Guard product detects prompt injection, jailbreaks, PII, and unsafe content using proprietary ML detectors trained on large adversarial datasets (including data from its Gandalf project). In September 2025 Lakera was acquired by Check Point Software (reported at ~$300M; deal value not officially confirmed) and now anchors Check Point's AI security center of excellence.
Lakera Guard is detection-first: it returns a flag/score and the calling application implements its own response logic. Its verdicts are ML-derived (a model_version field is exposed), making them probabilistic rather than deterministic — excellent for catching adversarial inputs, but not a reproducible, rule-attributable compliance control.
EVE CoreGuard is the opposite design point: a deterministic, fail-closed gate that returns the same verdict for the same input, attributes it to a named rule in a versioned regulatory pack, and emits a signed certificate an auditor can verify offline. Many teams run a security layer like Lakera and a deterministic governance gate like EVE CoreGuard.
Proprietary ML detectors trained on tens of millions of adversarial data points (including the Gandalf challenge), with strong published detection rates — a genuine strength for open-domain LLM threat coverage.
A SaaS API (and self-hosted Docker/Helm option) marketed as deployable in minutes with no model or prompt changes — easy to drop in front of an LLM app.
Covers PII detection, content moderation, agent/MCP and RAG security — breadth across the LLM attack surface that a deterministic compliance gate does not attempt.
Compared on the dimensions that distinguish a deterministic governance enforcement plane from Lakera Guard.
| Dimension | EVE CoreGuard | Lakera Guard |
|---|---|---|
| Primary purpose | Deterministic pre-execution governance & enforcement (the enforcement plane) | AI security — ML-based detection of prompt injection, jailbreak, PII, unsafe content |
| Enforcement timing | Pre-execution gate — decides ALLOW / BLOCK / MODIFY before the action runs | Detection-first; docs recommend screening after the LLM response, before delivery; the app implements response logic |
| Decision model | Deterministic rule evaluation — same input always yields the same verdict | Proprietary ML detectors (probabilistic; versioned model, not deterministic rules) |
| Zero-LLM enforcement verdict | ✓ Zero-LLM enforcement verdict (Layer A) | — ML/model-based detection |
| Fail-closed default | ✓ Fail-closed by default | — Application decides the action on a flag; Publicly documented capability not identified. for infra-failure behavior |
| Cryptographic decision certificate | ✓ Ed25519-signed decision certificate per verdict | — Publicly documented capability not identified. |
| Offline / replay verification | ✓ Offline + replay verification | — Publicly documented capability not identified. (ML model versions evolve) |
| Runtime attestation | ✓ Runtime attestation (attestation-bound execution authority) | — Publicly documented capability not identified. |
| Signed audit lineage | ✓ Signed audit lineage (signed audit bus + Merkle roots) | Logs + SIEM export (Splunk, Grafana); cryptographic tamper-evidence not publicly documented |
| Regulatory policy packs | ✓ Executable packs: ECOA/Reg B, FCRA, SR 11-7, HIPAA, EU AI Act, NIST AI RMF | Aligns to OWASP LLM Top 10, NIST, EU AI Act; not executable regulatory enforcement packs |
| Prompt-injection / jailbreak detection | Out of scope — governance, not threat detection | ✓ Core strength |
✓ = publicly documented · Partial = partial / configurable · — = "Publicly documented capability not identified."
Lakera answers "is this input or output an attack or a leak?" with a trained model. EVE CoreGuard answers "does this action comply with policy, and can I prove the decision?" with deterministic rules and a signed certificate. A security detector that returns a confidence score is the right tool for adversarial robustness; it is not, by construction, a reproducible control an examiner can replay. EVE CoreGuard is built to be exactly that — and is not a prompt-injection detector.
Lakera Guard runs proprietary ML detectors and returns flags/scores; results can vary as detector models are updated. EVE CoreGuard runs deterministic rule evaluation with no model in the verdict path — same input, same decision, every time.
Lakera produces logs and SIEM-exportable events. EVE CoreGuard produces a signed, replayable certificate that a third party can verify offline against the exact rule and policy version that decided the action.
Lakera screens inputs/outputs for threats. EVE CoreGuard gates the action against compliance policy. Run both: Lakera for adversarial security breadth, EVE CoreGuard for provable, deterministic compliance enforcement.
Choose Lakera (now Check Point) when your primary need is AI security: detecting prompt injection, jailbreaks, data exfiltration, and unsafe content across open-domain LLM and agent applications, with fast developer integration. Its ML detectors and adversarial dataset are a real strength for threat coverage that a deterministic compliance gate does not provide.
Choose EVE CoreGuard when you need a deterministic, provable compliance control rather than probabilistic threat detection: the same verdict for the same input, attributed to a named rule in a versioned regulatory pack, fail-closed, and emitted as a signed certificate you can verify offline and replay for an examiner.
Book a review and we will walk your use case through EVE CoreGuard — including a signed decision record you can verify offline. Pilot from $37,500; Enforcement from $150,000/yr.
Comparison based on publicly available product documentation as of June 2026; competitor capabilities evolve — verify current specifics with each vendor. Capabilities not found in public documentation are marked "Publicly documented capability not identified." Each product named is a trademark of its respective owner; this independent comparison is not affiliated with or endorsed by them. Related: All comparisons · Benchmark · EVE CoreGuard.