ActiveFence is now Alice

x

Implementing Safer AI, together

Your AI stack is only as safe as
its weakest integration.

WonderSuite connects natively with the AI infrastructure, platforms, and frameworks your teams already use adding adversarial testing, runtime guardrails, and production evaluation without replacing your stack.

Integrate with us

Integration

Microsoft Copilot Studio makes it easy to build agents. WonderFence makes it safe to run them. Set policies, enforce guardrails, and monitor every prompt and action in real time.

Partner + Implementation

OpenAI's base model safeguards are tuned for general use not your organisation's specific risk profile, compliance obligations, or abuse patterns. WonderSuite layers on top of your OpenAI deployment to add the control, evidence, and governance that enterprise AI actually demands.

Partner + Integration

Databricks Mosaic AI enables teams to build and orchestrate AI agents Alice extends this with AI safety guardrails for production. See how WonderFence integrates to prevent prompt injection, enforce policy, and maintain control across AI systems in real time.

Partner + Integration

As enterprises deploy AI agents on NVIDIA infrastructure at scale, safety can't be bolted on at the application layer alone. Alice and NVIDIA integrate at the infrastructure level — ensuring that the next generation of AI agents is secured from the ground up.

Partner

Alice contributes real-world adversarial intelligence to OWASP's AI and LLM security frameworks and WonderSuite's testing methodologies align with OWASP Top 10 for LLMs, giving security teams a shared standard language for AI risk.

Partner + Integration

For enterprises with strict data residency requirements, WonderSuite deploys inside your AWS environment no interaction data leaves your perimeter. Govern AI agents running on AWS infrastructure with the same rigour you apply to every other workload.

Partner

Alice red-teamed Amazon Nova pre-launch using hybrid adversarial testing combining automated scenario generation with expert human analysis. The engagement established a validated safety baseline before the model reached production customers.

Partner + Implementation

Parlant enables organisations to build and deploy open-source AI agents. The Alice integration adds real-time safety and security infrastructure on top of Parlant's open framework so teams get agent flexibility without sacrificing enterprise governance.

Partner + Implementation

Synthetic media is now a first-class AI risk. Reality Defender's deepfake and synthetic content detection is built directly into WonderFence giving enterprises real-time protection against AI-generated audio, video, image, and text manipulation at interaction time.

Partner + Implementation

Regulatory requirements for AI are evolving faster than most compliance teams can track. Alice and OpenPolicy work together to translate production risk signals into policy and proactively keep WonderSuite's governance frameworks aligned with the EU AI Act, ISO 42001, and NIST.

Partner + Implementation

Most Trust & Safety systems focus on stopping bad behavior not encouraging prosocial behavior. See how Alice and Modulate use real-time feedback, policy design, and AI to promote prosocial behavior and build healthier online communities.

Partner with Alice to shape the future of trusted AI

We work with AI platforms, infrastructure providers, implementation specialists, and policy organisations to help enterprises build, launch, and scale trusted AI systems.

Integrate with Us

FAQ: What buyers ask before they integrate.

Does WonderSuite work across multiple AI providers simultaneously?

Yes. WonderFence is model-agnostic and operates at the application layer, not inside any specific provider. It works across OpenAI, Anthropic, Llama, Databricks, and self-hosted models simultaneously. You get a single governance layer and unified policy enforcement across your entire AI portfolio, regardless of which models underpin each application.

We're concerned about latency. Will WonderSuite slow down our customer-facing financial AI?

WonderFence operates at under 100ms, purpose-built for high-traffic, latency-sensitive environments like payment agents and customer-facing banking applications. Protection doesn't come at the cost of customer experience.

What does it mean that WonderSuite is powered by Rabbit Hole?

Rabbit Hole™ is Alice's adversarial intelligence engine built on over a decade of global trust, safety, and security research, and informed by more than 10 billion analyzed deceptive, toxic, and manipulative samples. When you run WonderBuild or WonderCheck evaluations, the adversarial scenarios aren't synthetic they're derived from real-world misuse patterns. This is the difference between testing against how systems fail in theory versus how they actually get exploited in production.

How do I apply to become an ecosystem partner?

Alice partners with AI infrastructure, platform, implementation, and policy organisations. If you build AI systems and want to strengthen safety and governance or extend WonderSuite’s capabilities apply via “Integrate with Us.”