StartupNew York, NY · Est. 2023
Haize Labs
Red-teaming and robustness for AI agents.
Haize Labs stress-tests agents with automated adversarial attacks — prompt injection, tool misuse, unsafe actions — and ships an SDK for continuous safety evaluation in CI.
Pros & cons
Each point cites its sources and says what kind of evidence backs it. A second reviewer tried to refute every claim before it was kept.
Pros
- Independent testKnown for automated red-teaming research, and co-built a Red-Teaming Resistance Leaderboard with Hugging Face.[1]
- MetadataIts Verdict library for stacking LLM-as-a-judge calls is MIT-licensed and works without the hosted product.[1]
- Independent testPublished research showing that LLM defenses do not hold up against multi-turn human jailbreaks.[1]
Cons
- Independent testBeacon Software acquired the team on September 17, 2026 as an internal research group, and the announcement made no continuity commitment to existing customers.[1]
- MetadataVerdict has gone quiet: its last PyPI release is 0.2.7 from July 2025, its last commit is from November 2025, and it has about 350 stars.[1][2]
- Independent testThe hosted platform and attack-generation algorithms are closed, and the public get-haized jailbreak repo is an unlicensed subset.[1]
- Best for
- Teams that want research-grade adversarial testing ideas or the open Verdict judge framework.
- Watch out for
- Haize is no longer a standalone vendor, so export your red-team corpus and line up a replacement.
Assessed Oct 29, 2026 · adversarially reviewedSuggest a correction
People
Co-founder & CEO
Evals & benchmarks
Open roles (6)
- Vertical AI Lead
New York City / San Francisco · Posted Apr 2, 2026
Engineering - Software Engineer Intern
NYC · Posted May 15, 2025
Engineering - Product Slinger
New York / San Francisco · Posted Apr 11, 2025
Engineering - Staff+ Software Engineer, Backend and Infra
New York, NY · Posted Apr 11, 2025
Engineering - Applied Researcher
New York City / San Francisco · Posted Jul 3, 2024
Research
Neighbours in the directory
San Francisco, CA · Est. 2022
The de-facto standard toolkit for building LLM applications and agents.
LangChain makes the LangChain framework and LangGraph, the low-level orchestration standard for stateful, controllable agents, plus the LangSmith platform for tracing, evaluating and monitoring them in production.
~$75M raised
San Francisco, CA · Est. 2023
Enterprise-grade evals, data and AI gateway.
Braintrust turns prompts and agent traces into testable datasets and CI, plus an inference gateway — helping teams measure whether harness changes actually improve agent quality.
Berkeley, CA · Est. 2020
Observability and evaluation for AI — makers of OSS Phoenix.
Arize traces and evaluates LLM and agent systems in production; their open-source Phoenix tracer is a common instrumentation choice for agent harnesses that need to be debuggable.