Terms of Service
Effective date: July 30, 2026 (v2 — see Version history below for what changed since June 28, 2026)
ASURIQ provides AI verification, cognitive enhancement, model routing, memory, code quality, browser automation, and multi-model deliberation services. These products are operated by Silent Ampersand LLC. By using asuriq.dev, the ASURIQ browser extension, API, CLI, or MCP server, you agree to these terms.
Accounts. You must provide accurate information when creating an account. You are responsible for maintaining the security of your credentials.
Subscriptions and credits. Some products require a monthly subscription. Others consume credits per action. Credits are purchased through Stripe checkout. Pack credits do not expire while your account is active. Subscription credits reset monthly. All purchases are final. Refunds are issued automatically when a credit-consuming action fails.
AI outputs. ASURIQ products that generate content (WorkForce deliberations, Cognitive Stack analyses, Code Quality reviews) use third-party language models. Verification products (Verify badge, detail cards) check AI outputs against public databases but do not guarantee accuracy. All outputs may contain errors. You are responsible for evaluating and verifying all outputs before relying on them for decisions. ASURIQ is a quality layer, not a guarantee.
Browser extension. The ASURIQ browser extension reads AI responses on supported platforms (ChatGPT, Claude, Gemini, and others) to provide verification. The extension does not read pages outside of supported AI interfaces. It does not collect browsing history, passwords, or form data. You may uninstall the extension at any time.
Where ASURIQ operates. Verification is not limited to AI chat responses. ASURIQ also checks claims directly on web pages you visit, independent of any AI conversation, through the same browser extension. Both surfaces — chat and web pages — use the same verification engine, the same badge, and produce the same kind of certificate.
Claim extraction. Before a claim is verified, ASURIQ extracts it from the surrounding text. That extraction step runs through your own connected AI provider and API key — the same one you already use — not a separate model ASURIQ operates. ASURIQ’s servers receive the extracted claim for verification, not the full page or conversation it came from.
API keys (BYOK). When you provide your own API keys for model routing, those keys are encrypted at rest and used only to route your requests. We do not use your keys for any purpose other than executing your requested actions. You are responsible for the costs incurred through your API keys.
Corrections and user-filed evidence. If you submit a correction or counter-evidence through asuriq.dev/corrections, that submission is your own statement, not ASURIQ’s. ASURIQ records what you submit, weighs it against the existing evidence, and updates a verdict if the evidence changes — it does not author, endorse, or edit your submission on your behalf. Corrections are public and logged, not private support tickets.
Certificates. Every ASURIQ verdict resolves to a certificate at a permanent URL (asuriq.dev/v/[id]). You may link to, embed, or cite a certificate URL freely — that is what it is for. You may not alter, obscure, or misrepresent what a certificate says, and an embedded certificate badge must continue to link to its real, resolving certificate.
Service availability. ASURIQ depends on third-party AI providers (Anthropic, OpenAI, Google, DeepSeek) and database services (Semantic Scholar, CrossRef, PubMed, and others). These services may experience downtime, API changes, or rate limiting outside our control. We provide error handling and diagnostics but cannot guarantee uninterrupted service.
Acceptable use. You may not use ASURIQ to generate content that violates applicable law, infringes intellectual property rights, or harms others. We reserve the right to suspend accounts that violate these terms.
Limitation of liability. ASURIQ is provided as-is. We make no warranties regarding the accuracy, completeness, or suitability of AI-generated outputs. Our liability is limited to the amount you have paid in the prior 12 months.
Changes. We may update these terms. Continued use after changes constitutes acceptance.
Version history
v2 — July 30, 2026 (this version).Added five sections: where ASURIQ operates (web pages as well as AI chat), claim extraction running on your own connected AI model, corrections and user-filed evidence, certificate use, and ingredient-brand licensing (Tranche, DeDact, VIGIL). No existing commitment from v1 was altered — additions only.
v1 — June 28, 2026. Initial terms: accounts, subscriptions and credits, AI outputs, the browser extension, BYOK API keys, service availability, acceptable use, limitation of liability, and the right to change these terms.
Questions: hello@asuriq.dev