Responsible AI assistance

Smart Update can extract structured drafts; CV Clinic can evaluate a selected CV against user goals or a pasted role; Career Agent can rank jobs and prepare application material. Supporting classification, extraction, matching, and safety checks may also use configured AI providers.

AI output is a suggestion, not a verified fact or professional decision. Users must review factual accuracy, dates, qualifications, claims, recipients, attachments, and job requirements before publishing or sending anything.

updated.cv does not promise interviews, hiring outcomes, ATS acceptance, or that an AI recommendation is unbiased or complete.

Only data needed for the requested feature should be sent, which can include CV content, profile entries, job descriptions, prompts, and selected application context. Do not provide secrets, unnecessary sensitive information, or third-party personal data you lack permission to use.

AI providers process information under their applicable service terms, configurations, locations, and retention controls. Their handling can differ by provider and may change; the Privacy Policy identifies the provider categories used by updated.cv.

When the deployment enables Personal AI, a user can connect an OpenAI, Anthropic, Google Gemini, or xAI API key and select a compatible model reported by that provider. The key is encrypted into an HTTP-only browser-session cookie, is not saved in the updated.cv database, is never returned to the settings page, and expires or is removed when the user disconnects it.

Personal AI requests still pass through updated.cv's server so the requested workflow can be assembled and protected. The selected provider receives the CV, role, job, or application context needed for that request. Provider charges, limits, availability, retention, and account terms belong to the user, while updated.cv's privacy and security obligations continue to apply.

AI may hallucinate, omit context, reproduce bias, or produce unsuitable language. Users and connected agents must not fabricate qualifications, impersonate another person, discriminate unlawfully, manipulate employers, or automate abusive or deceptive applications.

API and MCP users choose permission scopes. Connected agents must follow user instructions, minimize access, preserve truthful data, use profile-specific edits when appropriate, and obtain the user’s review before consequential external action. Access can be revoked from Integrations.

Users may reject AI output, edit or delete generated content, disconnect agents, request account export or deletion, and contact support about an AI-assisted result or data concern.