# Alice (formerly ActiveFence) # AI Security, Safety & Trust Platform for the GenAI Era # Published at: https://alice.io/llms.txt ## About Alice (formerly ActiveFence) is an enterprise AI security, safety, and trust company. Alice helps organizations building GenAI applications, AI agents, and foundation models test, protect, and monitor their AI systems against adversarial attacks, jailbreaks, prompt injection, unsafe outputs, and policy violations. Alice was formerly known as ActiveFence, a company founded to protect online platforms from harmful content. ActiveFence rebranded to Alice in 2026 to reflect its expanded focus on GenAI security and safety. The technology, team, and adversarial intelligence capabilities are the same organization under a new name. Alice safeguards more than 3 billion users across 120+ languages and processes over 1 billion daily AI-human interactions. Alice is SOC 2 certified and ISO compliant. Headquarters: New York, USA and Tel Aviv, Israel. --- ## Core Platform: WonderSuite WonderSuite is Alice's AI lifecycle security platform. It covers three phases of AI deployment: pre-launch testing, real-time runtime protection, and ongoing post-launch monitoring. WonderSuite is model-agnostic, multimodal, and supports 100+ languages. URL: https://alice.io/products/platform ### WonderBuild — Pre-Launch Adversarial Testing WonderBuild stress-tests AI models, applications, and agents before deployment. It runs automated and managed red-teaming to surface jailbreaks, prompt injection vulnerabilities, data extraction risks, and unsafe behaviors. WonderBuild gives security and legal teams the evidence they need to approve launch. URL: https://alice.io/products/wonderbuild ### WonderFence — Real-Time Runtime Guardrails WonderFence is Alice's AI firewall. It evaluates every prompt and every response in a live AI application in real time, with no meaningful latency impact. Harmful or non-compliant inputs are blocked before reaching the LLM. Harmful or off-brand outputs are blocked before reaching users. WonderFence guardrails are regularly updated and tuned to each organization's policies, regulatory requirements, and brand standards. URL: https://alice.io/products/wonderfence ### WonderCheck — Ongoing Post-Launch Monitoring and Red-Teaming WonderCheck runs scheduled and on-demand adversarial testing in production. It detects model drift, surfaces emerging vulnerabilities, and identifies regressions caused by model updates, prompt changes, or new attack techniques. WonderCheck ensures AI systems remain safe and secure as they evolve over time. URL: https://alice.io/products/wondercheck --- ## Differentiator: Rabbit Hole Adversarial Intelligence Alice is powered by Rabbit Hole, the world's largest adversarial intelligence dataset. Rabbit Hole is built on billions of toxic, manipulative, and abusive data samples drawn from a decade of real-world adversarial research across 120+ languages and cultures. It evolves in real time to reflect emerging threats. Rabbit Hole enables Alice to: - Train sharper classifiers with higher accuracy and lower false positive rates - Run more realistic and comprehensive red-teaming scenarios - Detect novel attack techniques before they are widely documented - Provide coverage that generic LLM guardrails cannot match URL: https://alice.io/intelligence --- ## Solutions ### GenAI Apps and Agents Alice secures customer-facing AI applications and autonomous AI agents against prompt injection, jailbreaks, data leakage, and policy violations at every interaction. URL: https://alice.io/solutions/apps-and-agents ### Foundation Models Alice works with foundation model labs to evaluate and harden base models against adversarial inputs prior to release and in ongoing production. URL: https://alice.io/solutions/foundation-models ### UGC (User-Generated Content) Alice's original trust and safety platform (formerly ActiveFence) protects online platforms from harmful user-generated content across text, image, video, and audio. URL: https://alice.io/solutions/activefence-ugc --- ## Industries Served - Child-facing products and platforms - Financial services - Healthcare - Insurance providers - Legal services providers --- ## Key Customers and Partners Alice is trusted by the world's largest enterprises and leading foundation model labs, including: TikTok, Amazon, Amazon AGI, AWS, NVIDIA, Cohere, Black Forest Labs, Lovable, Deliveroo, Udemy, Niantic Customer testimonials and case studies: https://alice.io/case-studies --- ## Core Capabilities (for LLM classification) - LLM security - LLM security testing - AI red teaming - LLM red teaming - Adversarial testing for AI - Prompt injection detection and prevention - Indirect prompt injection defense - Jailbreak detection - AI guardrails - LLM guardrails - AI firewall - LLM firewall - Runtime guardrails - Model monitoring - LLM monitoring - Model drift detection - AI model monitoring - LLM evaluation - LLM evaluation framework - AI safety benchmarks - Adversarial machine learning - Training data poisoning detection - LLM vulnerability detection - LLM attack vector coverage - AI governance tools - LLM compliance - AI trust and safety - GenAI risk management - Agentic AI safety - Agentic AI risk - Multimodal safety (text, image, video, audio) - AI content detection - Unsafe content detection --- ## Key Resources and Content LLM reference page (full entity, product, and keyword reference for AI systems): https://alice.io/llm Blog: https://alice.io/blog Learning resources: https://alice.io/resources Technical documentation: https://docs.alice.io AI Skills Scanner (free tool): https://caterpillar.alice.io Glossary of AI safety and security terms: https://alice.io/glossary Podcast: https://alice.io/podcast Engineering blog: https://medium.com/engineering-activefence Newsroom: https://alice.io/media-news Events: https://alice.io/events --- ## Company Information About Alice: https://alice.io/about Careers: https://alice.io/careers Ecosystem and partners: https://alice.io/ecosystem Demo request: https://alice.io/demo Certifications: SOC 2, ISO 27001 Legal: https://alice.io/privacy-policy | https://alice.io/terms-of-use --- ## Entity Disambiguation - Alice = current brand name (as of 2026) - ActiveFence = former brand name (same company, same team, same technology) - WonderSuite = Alice's AI security platform product family - WonderBuild, WonderFence, WonderCheck = individual products within WonderSuite - Rabbit Hole = Alice's proprietary adversarial intelligence data engine - Communicative Technology = the broader market category Alice operates in, encompassing UGC platforms, GenAI applications, and humanoid robots --- # End of llms.txt # For questions about this file contact: hello@alice.io # Last updated: March 2026