userfront.com

What it is

Userfront currently positions AI Guardrails as an “AI Portal for Professionals” and notes that Userfront AI has become MultiModel. It is not a single large-model app, but an enterprise AI portal and governance layer: employees access company-approved AI assistants through a unified hub, while IT manages access control, policy enforcement, logging, and compliance.

Core capabilities

In terms of model coverage, the page explicitly lists ChatGPT 4o, o1, 4o-mini, o1-mini, Microsoft Copilot Chat, Azure OpenAI, as well as ChatGPT, Microsoft 365 Copilot, Anthropic Claude, and Google Gemini under the Multi-Assistant offering. The page also shows entry points for Grok, Sonar, Perplexity, Custom Assistants, RAG Models, and others, though details are limited. Its core value is not generation itself, but Guardrails: sensitive data is redacted in real time before prompts are sent; access can be controlled by user, group, policy, assistant, and model; and user prompt logs are retained. The single-assistant plan mentions 24-month log retention, while the multi-assistant plan supports custom retention periods.

Compliance, integrations, and privacy

Userfront highlights SOC 2 Type 2 coverage across all 5 Trust Services Criteria, with annual audits by Ernst & Young and continuous monitoring by Drata. It also mentions monthly third-party vulnerability scans and penetration tests, automated security reports, activity logs, and more. For integrations, it supports mainstream SSO and identity systems such as Microsoft Entra ID, Okta, LDAP, AD, and SAML, making it suitable for enterprise identity environments. On privacy, the page states that user inputs in the ChatGPT offering are not used for OpenAI training, and that Copilot inputs are not used for Microsoft or OpenAI training.

Pricing and trial

AI Guardrails does not publish pricing. Pricing is based on seat count and deployment location, with support for cloud, hybrid, and on-premises deployment. All AI plans require contacting sales. The FAQ says evaluation is possible, but sales must be contacted to start configuration; no clear free tier was found.

Pros, cons, and who it is for

Its strengths are comprehensive enterprise governance capabilities, covering multiple models, multiple assistants, SSO, data redaction, auditing, and compliance. It is especially suitable for highly regulated industries such as finance, insurance, and healthcare. The drawbacks are opaque pricing and limited self-service information for individuals and small teams. Chinese-language support, China deployment, and payment methods are also not disclosed. Another limitation is that it is primarily an AI management and security layer, so answer quality still depends on the connected third-party models.

Access from China

The main content does not mention access from mainland China, network availability, or RMB payment. Given its reliance on services such as OpenAI, Anthropic, Google, and Microsoft, actual use in China may be affected by network and compliance restrictions. Before procurement, confirm accessibility, data residency, payment methods, and alternative model options.