Alice
Founded in 2024, Alice provides end-to-end AI safety and security tools - stress-testing, runtime guardrails, red-teaming and drift detection - for AI models, applications and agents.
Alice is an AI safety and security company founded in 2024. It provides end-to-end solutions spanning the AI lifecycle, from pre-launch testing to production monitoring, for AI models, applications and agents. The company was established to address gaps left by legacy trust and safety tools in the AI era, and it operates globally.
Its flagship platform, WonderSuite, combines several components. WonderBuild handles pre-launch stress-testing of AI models. WonderFence provides dynamic runtime guardrails for AI systems in production. WonderCheck delivers ongoing automated red-teaming and model drift detection. Underpinning these is Rabbit Hole, an intelligence engine drawing on a decade of real-world adversarial research, which processes billions of toxic data samples to power classifiers and safety evaluations.
Technical work at the company spans AI safety and security, generative AI, prompt injection and jailbreak detection, machine learning classifiers, adversarial research and AI model evaluation, with coverage of more than 120 languages. Alice reports that it is trusted by 8 of the 10 leading frontier AI labs, and that its systems safeguard more than half of the world's online experiences, protecting over 3 billion users and securing one billion daily AI-human interactions.
Customers and use cases extend across financial services, healthcare, insurance, child-facing platforms and AI/ML development. The company describes its mission as enabling organisations to advance unafraid, with a focus on testing, protecting and monitoring AI applications and agents.
