Voice & AI Service Terms

Effective Date: June 2, 2026

1. AI Identification

Inflowence provides automated Voice Assistants. Customer is solely responsible for determining, and complying with, applicable law in each jurisdiction where its Voice Assistants initiate or receive calls.

Jurisdictions that require AI disclosure. A growing number of states — including, at present, California, Utah, and Colorado — require clear and conspicuous disclosure that a communication is automated or AI-generated, and these laws are changing frequently. Federal law (the Telephone Consumer Protection Act and the FCC's 2024 AI-voice ruling) imposes identification and consent requirements on outbound artificial-or-prerecorded-voice calls, but by its terms does not reach inbound calls answered on a called party's behalf; the inbound disclosures in Section 2 below rest on state law, not on it. Customer is responsible for determining, and complying with, the requirements applicable in each jurisdiction where its callers are located.

Where any such law applies, Customer must configure the Voice Assistant to make a clear and conspicuous disclosure, in the Voice Assistant's first utterance and before any request for information, that the caller is speaking with an "AI Assistant," "Automated System," "AI Receptionist," or substantially similar language. The word "virtual" alone does not satisfy the clear-and-conspicuous requirement, because "virtual receptionist" is a well-established industry term for a remotely-located human receptionist and a reasonable consumer may understand it that way.

Jurisdictions that do not require AI disclosure. In jurisdictions without an applicable bot-disclosure statute - including, as of the Effective Date above, Arizona and most other U.S. states for inbound customer-service calls - such disclosure is not legally required, and Customer may choose whether its Voice Assistant proactively identifies itself as AI. Inflowence recommends disclosure as a best practice even where not required, because undisclosed AI interactions can still expose Customer to claims under general consumer-protection laws, including the Arizona Consumer Fraud Act (A.R.S. § 44-1521 et seq.), Section 5 of the FTC Act (15 U.S.C. § 45), and analogous state unfair-and-deceptive-acts-and-practices statutes, where a reasonable consumer would have believed they were speaking with a human. Customer assumes all risk of a non-disclosure decision.

Prohibited in all jurisdictions, regardless of disclosure choice. Customer shall not configure, prompt, or otherwise cause a Voice Assistant to:

impersonate a specific natural person;
impersonate, or falsely suggest affiliation with, any federal, state, tribal, or local government agency, court, or law-enforcement body;
impersonate, or falsely suggest affiliation with, any actual business, brand, financial institution, healthcare provider, insurer, utility, or other real entity without that entity's express written authorization; or
affirmatively represent to an End-User that the Voice Assistant is a human being, or deny being AI in response to a direct question from the End-User. Silence as to the Voice Assistant's automated nature, in a jurisdiction where proactive disclosure is not required, is not an affirmative representation and is permitted; a spoken denial of AI status is not.

This obligation reflects, without limitation, the FTC Rule on Impersonation of Government and Businesses (16 CFR Part 461), applicable state anti-impersonation statutes, and general UDAP principles.

Customer responsibility. Customer is solely responsible for its prompts, agent configurations, Voice Assistant self-identification choices, and deployments in every jurisdiction in which it operates. Inflowence is not liable for Customer's configuration choices or for Customer's failure to assess or comply with applicable law. Customer's indemnification obligations under the Terms of Service apply to any claim arising from a disclosure or non-disclosure decision.

Certain jurisdictions require the consent of all parties to a recorded call. Inflowence obtains that consent from the caller, on every call, in every state. Two disclosures do this, and Inflowence configures both (see MSA §6.2(a)):

1. Recording announcement — played before the call is answered, and before any recording begins. It must state that the call is being recorded and transcribed. Default wording:

"This call is recorded and transcribed for quality purposes."

A caller who continues the call after hearing this has consented to being recorded. Notice comes first; recording follows.

2. AI and answering-service disclosure — the first thing the Voice Agent says. It must name the business, declare that the caller is speaking with an artificial intelligence rather than a person, and declare that a third-party answering service is handling the call on the business's behalf. Default wording:

"You've reached [Business]. I'm [Agent Name], their AI answering service. How can I help?"

Wording may vary; the disclosures may not. Customer may request alternative phrasing, and Inflowence will adopt it if it still satisfies every required element above. For example, "This call is recorded for quality purposes. Connecting you now." is fine. "Thanks for calling!" is not, and Inflowence will refuse it — including over Customer's instruction. Inflowence, not Customer, is responsible for disclosure content (MSA §6.2), and every candidate disclosure is validated in software before it is deployed.

No consent gate. The Voice Agent does not ask the caller to say "yes" before proceeding. A recording, once started, cannot be stopped mid-call on our telephony platform — so asking a question we could not act on would be worse than not asking. Notice before recording, followed by the caller's continued participation, is the consent.

If a caller objects to being recorded, the Voice Agent acknowledges the objection, offers a return call on the number they called from, and ends the call. The audio recording of that call is deleted (MSA §6.2(d)). No live-transfer option is offered, and none is represented to exist — the Voice Agent answers precisely because no one else is available.

Two-party / all-party consent states. Jurisdictions that require consent from all parties to a recorded call include, without limitation, California (Penal Code §§ 631, 632, 632.7), Florida (Fla. Stat. § 934.03), Illinois, Maryland, Massachusetts, Montana, New Hampshire, Pennsylvania, and Washington. California is the strictest: Penal Code § 632.7 reaches any call involving a cellular or cordless telephone and has no "confidential communication" element.

Customer responsibility. Inflowence configures and is responsible for the disclosures above. Customer must not disable, shorten, or alter them, and must not configure a Voice Agent to answer without them.

3. Biometric Data Disclaimer

Inflowence and its sub-processors (GoHighLevel and the providers GoHighLevel engages for telephony and transcription) process voice data for transcription and intent recognition only. Inflowence does not use voice data to identify specific individuals or for biometric "fingerprinting." Customer is prohibited from using the Service to collect or process biometric identifiers.

Customer biometric prohibition. Customer shall not use the Service to create voiceprints or voice signatures, perform biometric authentication, or maintain databases of biometric voice data, including any use that would trigger state biometric privacy laws such as the Illinois Biometric Information Privacy Act (BIPA, 740 ILCS 14), the Texas Capture or Use of Biometric Identifier Act (CUBI, Bus. & Com. Code § 503.001), Washington RCW 19.375, or the California Consumer Privacy Act's sensitive personal information provisions. Because Inflowence does not create biometric identifiers through transcription and intent recognition, these laws do not apply to Inflowence's processing; they may apply to Customer's configuration choices.

4. AI "Hallucinations" & Liability

Inflowence provides the Voice Agent "as-is." Customer is responsible for reviewing the logic and responses of their AI agents. Inflowence is not liable for any commitments, contracts, or misinformation provided by an AI Agent to an End-User.

Indemnification. Customer shall indemnify and hold harmless Inflowence from any claims, damages, or liabilities arising from the Voice Agent's responses, commitments made by the Voice Agent, misinformation conveyed by the Voice Agent, or Customer's failure to properly configure or monitor its Voice Agents.

5. Data Processing, Retention & Sub-Processors

Call recordings, transcripts, and associated metadata are retained in accordance with the retention periods configured on Customer's account and the Inflowence Privacy Policy. Voice data is processed by GoHighLevel (GHL) for call routing, transcription, and management, together with the telephony and AI-model providers GHL engages as its own subprocessors. A current list of sub-processors is available at Subprocessors.

Contact Us

For questions about voice and AI compliance:

Email: compliance@inflowence.ai

By using Inflowence's Voice & AI features, you acknowledge that you have read, understood, and agree to this policy.