Echo One /

Acceptable use and AI responsibilities

Use powerful assistance with clear responsibility.

Rules for the Echo One website, demonstrations and clinical platform. Clinical use requires an agreed service scope and the relevant customer agreement.

Last updated · 25 September 2026

01

Stay within the agreed purpose.

Use each feature only for its documented and contracted purpose, within your professional authority and the permissions granted to you. A demonstration, experimental feature or draft does not establish suitability for clinical use. Do not use Echo One as an autonomous clinician, an emergency service or a substitute for required monitoring.

02

Review before relying on AI output.

Echo remains responsible for its contracted service and applicable legal duties. Clinician review is a safeguard and professional responsibility, not a waiver of the supplier's obligations.

  • Check the patient, source information and date before reviewing an output.
  • Look for omissions, invented details, misleading summaries and incorrect codes as well as obvious errors.
  • Resolve uncertainty, edit or reject the draft, and use the required approval or signature workflow before it enters clinical use or is shared.
  • Verify medicine identities, doses and clinical decisions using appropriate clinical sources and professional judgement.
  • A citation, confidence score or valid data format does not establish that a clinical claim is correct.
03

Respect patient choices and authority.

Explain recording and AI assistance before use and obtain the notices, permissions or consent required for the activity. Follow the service's process for refusal, withdrawal, capacity and authorised representatives. Do not record someone covertly or treat website registration as patient consent. Use only records and systems you are authorised to access.

04

Protect information and accounts.

Keep credentials personal, use required authentication controls, grant only appropriate access and promptly report suspected compromise. Use approved clinical channels for health information. Do not put it into marketing forms, public feedback, demonstration environments or unapproved third-party AI tools.

05

Do not bypass the safeguards.

  • Do not attempt to access another practice's or person's information without authority.
  • Do not defeat human-review requirements, falsify signatures or conceal changes to a record.
  • Do not manipulate prompts, documents or tool responses to obtain prohibited actions or disclose confidential information.
  • Do not introduce malware, disrupt the service, evade agreed usage limits or perform unauthorised security testing.
  • Do not send unlawful messages, scrape personal data, impersonate people or use the service for harassment, fraud or discrimination.
  • Do not represent demonstrations or generated content as verified clinical evidence or claim authorisations that have not been established.
06

Report problems in context.

If output appears unsafe, stop relying on it, arrange appropriate clinical review and use the agreed incident process. Contact security@echo-health.ai for security or privacy concerns. Describe the feature and issue without emailing patient records; request an approved secure channel when more detail is necessary. Urgent care must use the appropriate healthcare pathway.

07

Proportionate action and review.

The proposed service agreement allows proportionate restrictions where needed to address a material security, legal or service-integrity risk. Where practicable, explain the reason and affected scope, allow correction and review the restriction promptly. Suspension must consider continuity of care and lawful access to records; these rules do not authorise punitive deletion of clinical data.

08

Related documents.

How our AI is governed

The architecture behind proposals, checks and accountable decisions.

Explore the safeguards

Terms of service

The proposed allocation of supplier and customer responsibilities.

Read service terms