Sference Acceptable Use Policy

This Acceptable Use Policy ("AUP") applies to everyone who uses Sference's services, including customers, their users and the end users of products built on Sference. It forms part of the Sference Terms of Service. Customers are responsible for their end users' compliance.

1. Unlawful and harmful use

Do not use the Services to:

• (a) break any applicable law or regulation, or help anyone else to do so;
• (b) generate, solicit or distribute child sexual abuse material or any content that sexualises minors;
• (c) threaten, incite or promote violence, terrorism or serious self-harm, or harass or abuse individuals;
• (d) create or distribute malware, or plan or carry out attacks on systems, networks or data that you are not authorised to test;
• (e) commit fraud, phishing or scams, send spam, or deceive people in a way that is unlawful or causes harm, including by impersonating a person or organisation without consent;
• (f) infringe or misappropriate anyone's intellectual property, privacy or other rights;
• (g) process personal data unlawfully, or submit special-category or criminal-offence data without the safeguards agreed with Sference in writing;
• (h) carry out any practice prohibited by Article 5 of the EU AI Act, including social scoring, untargeted scraping of facial images to build recognition databases, and manipulative or exploitative techniques that cause significant harm.

2. High-risk use

Do not use the Services:

• (a) as the sole basis for decisions that have legal or similarly significant effects on individuals (for example in employment, credit, insurance, housing, education, healthcare, migration or law enforcement) without meaningful human review;
• (b) in systems where failure or inaccuracy could cause death, personal injury or serious physical or environmental damage, including medical devices and the operation of critical infrastructure, unless agreed with Sference in writing;
• (c) to give the impression that AI-generated content was produced by a human where the law requires disclosure.

You are responsible for your own obligations under laws regulating AI, including the EU AI Act, in respect of the systems you build or deploy.

3. Model licences

Models in the catalogue are made available under their own licences and usage policies. Your use of a model must comply with the licence and usage policy that applies to it.

4. Sanctions and export control

Do not use the Services if you are subject to UK, EU or US sanctions, or located in a country or territory subject to comprehensive sanctions, and do not use them in breach of export-control laws.

5. Protecting the platform

Do not:

• (a) circumvent rate limits, quotas, usage controls or safety measures, or create multiple accounts to do so;
• (b) share, sell or transfer API keys or account access;
• (c) resell or sublicense access to the Services as a standalone inference or model-access service without a written reseller agreement with Sference;
• (d) probe, scan or test the vulnerability of the Services, or access other customers' data or accounts, without Sference's written permission;
• (e) reverse engineer the Services, or attempt to extract model weights or proprietary system components;
• (f) interfere with or place unreasonable load on the Services.

6. Enforcement

Sference may investigate suspected breaches of this AUP, including by automated means, and may suspend or restrict access, remove content, terminate accounts and report unlawful activity to the competent authorities, as set out in the Terms of Service. Sference may act without notice where it considers this necessary to prevent harm or comply with law.

7. Reporting

Report suspected abuse of the Services or security vulnerabilities to [email protected].

8. Changes

Sference may update this AUP at any time by publishing a new version on this page.