AI Terms

These AI Terms supplement the Terms of Service whenever you use features that generate, classify, summarize, retrieve, recommend, speak, transcribe, plan, or propose actions using machine learning or artificial intelligence (AI Features). They apply to typed, voice, workflow, and agentic interfaces. If these AI Terms conflict with the Terms, these AI Terms control only for AI Features.

You retain your rights in prompts, files, instructions, recordings, workspace context, and other material supplied to an AI Feature (Input). You represent that you have the rights and lawful basis needed to submit Input and authorize its processing.

Corporate Suite constructs bounded context from records the requesting user is authorized to access. Customers remain responsible for workspace permissions, connector scopes, data minimization, and deciding which information may be included in AI workflows.

Subject to applicable law and third-party rights, as between TWC and Customer, Customer may use output generated for Customer. TWC assigns to Customer any rights TWC may have in that output. This does not grant rights in the Service, models, system instructions, or third-party material.

AI output may not be unique, and other users may receive similar output. Output can contain errors or material that requires a license or attribution. You are responsible for evaluating and clearing output before publication or commercial use.

AI Features are probabilistic. Output may be false, incomplete, biased, offensive, outdated, or inconsistent. Do not treat output as a factual source or as legal, medical, financial, employment, safety, compliance, or other professional advice. Verify material facts, calculations, citations, permissions, and consequences with appropriate sources and qualified people.

A human must review and affirmatively approve a proposed action before it deletes or irreversibly changes data; communicates outside the workspace; incurs a charge or transfers value; changes access or identity; binds a person or organization; publishes content; or materially affects legal rights.

Customers must not disable, work around, or falsely satisfy an approval step. A separately contracted automation may use bounded pre-authorization only when it defines scope, limits, responsible owner, monitoring, revocation, and exception handling. TWC may require re-approval when context or risk changes.

AI Features may route Input and authorized context to configured model, search, speech, or tool providers identified on our Subprocessors page or in an enterprise agreement. Provider and model selection may change for performance, availability, safety, or capability. We use contractual and technical controls appropriate to the service, but third-party processing remains subject to its agreed service terms.

We do not use Customer Content to train general-purpose AI models unless Customer expressly opts in under a separate agreement that identifies the data and purpose. We may use de-identified, aggregated service telemetry and customer feedback to evaluate reliability and improve the Service where that data does not identify Customer or individuals.

We may perform limited human review of content when requested for support, needed to investigate abuse or security, legally required, or expressly authorized for an evaluation program. Access is restricted and logged according to our security program.

You must follow the Acceptable Use Policy. You may not use AI Features for unlawful surveillance; prohibited discrimination; impersonation or material deception; non-consensual sexual content; exploitation of children; malicious code or credential theft; evasion of safeguards; or solely automated high-impact decisions about individuals. Customers deploying AI-generated content to the public must make disclosures where law or context requires them.

Start a voice feature only when you are authorized to record or transmit the participants. You are responsible for notices and consent required by wiretap, call-recording, biometric, employment, and privacy laws. Audio, transcripts, and derived output may be processed by a configured voice provider and retained according to the feature, customer settings, and our Privacy Policy.

We may apply automated and human safety checks, rate limits, audit lineage, and provider usage measurement. We may block an Input, withhold an Output, require additional approval, or suspend an AI Feature when reasonably necessary to address abuse, security, legal risk, provider restrictions, or harm. These controls do not replace Customer's own governance.

AI FEATURES AND OUTPUT ARE PROVIDED “AS IS.” TWC DOES NOT WARRANT ACCURACY, UNIQUENESS, NON-INFRINGEMENT, OR FITNESS FOR A PARTICULAR DECISION OR WORKFLOW. The disclaimers and liability limits in the Terms apply. Nothing in these AI Terms limits rights that cannot lawfully be waived.