Enterprise AIs versus Public AIs
Enterprise AIs are AIs that are not publicly available and are licensed by companies to be used within companies across the companies proprietary data. They usually limit the amount of data being sent back to the AI company for training purposes, as it is proprietary data (trademarks, copyright, trade secrets), or for privacy reasons. Works created by licensed AIs within a licensed company, the copyright would remain with the company.
Public AIs are open to anyone AIs (free or subscription), these AIs are more likely to use any inputs for training the AI (check your terms of service), and it is less likely that outputs have any copyright protection.
AIs contained within traditional software.
Many traditional software packages now include an AI within their system, for example Adobe has Firefly. If you have a licence to use the product, then you can use the AI. Check your terms of service for concerns regarding whether inputs are used for training, and whether outputs can be protected by copyright.
Copyright ownership of prompts
It is recognised that you can own the copyright of the prompts that you enter into an AI. Standard copyright rules apply.
Copyright ownership of outputs
Ownership of copyright works requires sufficient human input. For AIs, this area is controversial at the moment. It is generally thought that there is insufficient human input for an AI's output to be protected by copyright, but this may change in the future.
Do:
Don’t:
AI has the potential to assist people in many capacities. Despite this, AI raises issues and risks for Copyright, Intellectual Property (IP), and Indigenous Cultural and Intellectual Property (ICIP).
Do not upload Indigenous Cultural and Intellectual Property materials (ICIP) into any AI without first obtaining the free and informed consent of cultural rights holders.
Please see the following posts for more insights.