Third-Party Restrictions

Thank you for using Scale Donovan! This document contains information about Third-Party Restrictions governing the use of Third-Party Offerings made available in Donovan. Capitalized terms used in this document but not defined below are defined in Scale’s Master Services Agreement (“MSA”) available at scale.com/legal/msa.

Scale does not endorse or make representations or warranties about Third-Party Offerings, and use of Third-Party Offerings is at your own risk. You are responsible for ensuring your use of Third-Party Offerings complies with any Third-Party Restrictions.

While Scale has sought to provide complete and accurate information for each Third-Party Offering, Scale does not represent or warrant that the licensing information and use case restrictions provided herein are correct or error-free. You should investigate the identified Third-Party Offerings and governing Third-Party Restrictions to confirm the accuracy of the information provided herein. You are encouraged to notify Scale of any inaccurate information or errors found in this document.

OpenAI

OpenAI Terms of Use governs the use of:

GPT-3 by OpenAI
DALL-E by OpenAI

OpenAI’s usage policies center on the safe and responsible use of their API. As of November 9, 2022, OpenAI no longer requires users to register their applications with OpenAI. Instead, OpenAI will use a “combination of automated and manual methods to monitor for policy violations”. You are encouraged to reach out to OpenAI’s safety specialists at [email protected] if you have questions about whether your application meets OpenAI’s usage policies.

OpenAI’s usage policy prohibits building products targeting the following use cases:

Illegal or harmful industries;
Misuse of personal data;
Promoting dishonesty;
Deceiving or manipulating others;
Trying to influence politics.

It also prohibits you or your application’s end users from generating the following types of content:

Hate
Harassment
Violence
Self-harm
Sexual
Political
Spam
Deception
Malware

In the event of high-stakes use cases that do not fall under the list of prohibited use cases above, OpenAI mandates that you thoroughly test your models for accuracy, and be transparent with your users about limitations. Additionally, you must ensure your team has domain expertise and understands/follows relevant laws.

OpenAI’s Terms of Use also restrict:

Releasing a foundation model, large-scale model or application that directly competes with OpenAI;
Reverse engineering or attempting to discover the source code, models, algorithms and systems of OpenAI’s “LLMs”.

For more information about OpenAI’s usage policies, please view their Terms of Use and Usage Policies.