Giskard vs WitnessAI

Giskard is a Paris-based company founded in 2021 by Alex Combessie, Jean-Marie John-Mathews and Andrei Avtomonov. Combessie and John-Mathews are now co-CEOs, and Giskard says its co-founders previously built AI systems at Dataiku and Thales. It sells software and services for testing and securing LLM agents. Its first funding round, €1.5 million led by Elaia with Bessemer Venture Partners, was announced in December 2022. Giskard says it also received a €3 million strategic investment through the European Commission's EIC Accelerator in 2023 and a €3 million Bpifrance grant in 2024 for LLM evaluation research with Mistral AI, Artefact, INA and BnF. CORDIS lists a €2.5 million EU contribution to its EIC project, which ran from September 2023 to August 2025. Giskard Hub is the paid platform. Continuous red teaming runs multi-turn attacks that adapt to the agent's answers, generated from the company's own documents and knowledge bases. It tests for hallucinations, stereotypes and discrimination, harmful content, disclosure of personal information and prompt injection. Testing is black-box: the agent only needs an API endpoint, and coverage is limited to text-to-text conversational agents. The Hub also schedules agent evaluations and manages test datasets, custom checks and human review, with alerts, SSO, role-based access and audit logs. It runs as SaaS, in a private cloud or on premises, with EU or US data processing. Giskard Guards, launched in April 2026, is a runtime guardrail layer that runs in the customer's infrastructure. It inspects prompts, responses, tool calls and multi-step reasoning, and its policies are written as code in OPA/Rego. Giskard also sells AI security audits that end in a go/no-go report, and it is offering a free trial of a new Agentic Red Teamer to a limited number of users. The open-source library is Apache 2.0 licensed. Version 3, a rewrite for multi-turn agent testing, became stable in August 2026. Version 2 is no longer actively maintained, and tabular models have no v3 equivalent. Giskard also publishes Phare, an LLM safety benchmark launched in February 2025 with Google DeepMind as research partner. Logos on its site include BNP Paribas, AXA, BPCE, Société Générale, Crédit Agricole, Michelin, Decathlon, L'Oréal and Doctolib. **AI compliance coverage** Guards includes an EU AI Act policy pack for prohibited practices such as social scoring and real-time biometric identification, and Giskard says Guards covers OWASP LLM risks. It doesn't map its products to the NIST AI RMF or ISO 42001. Its trust center, run on Bastion, lists GDPR and a SOC 2 Type 2 report and shows HIPAA as in progress. **Who it's for** Enterprises, especially in European banking and insurance, that run customer-facing LLM agents and want testing and runtime controls they can deploy on premises.

WitnessAI is an AI security and governance platform for enterprise AI use by employees and by AI agents. It was founded in 2023 and incubated by Ballistic Ventures. CEO Rick Caccia, previously SVP of marketing at Palo Alto Networks, is the only co-founder the company names. It is based in Mountain View, California, and has raised about $85.5 million in two announced rounds: $27.5 million co-led by GV and Ballistic Ventures (May 2024) and $58 million in strategic funding led by Sound Ventures (January 2026), with Fin Capital, Qualcomm Ventures, Samsung Ventures and Forgepoint Capital Partners taking part. The platform has three modules. Observe finds the AI apps, agents and MCP servers in use, drawing on a catalog of thousands of AI applications, and classifies conversations by intent. Control sets policies by department, role, intent or workforce type, routes prompts to approved or cheaper models, limits agents to approved MCP servers and tools, redacts sensitive data in real time and keeps audit trails. Protect (Witness Protect, which WitnessAI calls "the enterprise AI firewall") screens prompts and responses for prompt injection, jailbreaks and harmful output and tokenizes PII, credentials and secrets. Witness Attack runs automated red teaming against models before deployment. WitnessAI says it sees AI activity across the network without browser extensions or endpoint clients, including native apps such as Windows 11 Copilot and Office 365, and it integrates with F5 BIG-IP SSL Orchestrator for encrypted AI traffic. Each customer gets a single-tenant instance with customer-controlled encryption. In January 2026 it launched WitnessAI Agentic Security, which covers coding agents such as Claude Code, Cursor and Codex and records the user, agent, tool and rule for every blocked tool call. In July 2026 it added AI FinOps features: spend estimates for unsanctioned AI apps and agents, metering of sanctioned token use, an ROI dashboard and cost-based model routing. **AI compliance coverage** The compliance page offers logging of every AI app and interaction, policy enforcement and masking of regulated data, but names no AI framework. WitnessAI's AWS Marketplace listing says its audit trails support compliance reporting for the EU AI Act, ISO/IEC 42001, the NIST AI RMF and PCI DSS 4.0.1. We found no control mappings for any of them. WitnessAI announced a SOC 2 Type II attestation in October 2025; it has no public trust center. **Who it's for** Large enterprises that want to see and govern how employees and agents use public and internal AI tools, and to add runtime protection to their own AI apps. WitnessAI says its customers include financial services firms, utilities, airlines, automakers, retailers and telecoms. InComm Payments is the only customer it names on its site.