According to Manus' Privacy Policy, Manus collects and processes a wide range of user personal information, including:
Although it is worth noting that Manus does not use personal data for the purpose of training AI models, it may be transmitted to third-party AI providers. Therefore, in class we need:
<aside> 📌
However, Manus’s data practices lack transparency, the company does not explain the policies of who trained its models, what datasets were used, etc. Because the official documents offer no clear explanation, we should be especially mindful of how Manus handles information and influences our teaching practices.
</aside>
Since Manus relies on a third-party sources that is trained based on large-scale internet data, many of its outputs (such as chemical diagrams or explanatory text) may indirectly reference information from publicly available scientific publications or educational resources.
However, we are not certain whether the materials included in the underlying training data of Manus were collected without the consent of individuals or proper attribution. This practice raises issues related to copyright rights of AI. Even if the output of the AI appears "novel", it could still be a fusion of existing human works. If users present the content generated by the AI as completely original content without indicating its possible derivative works. And then without realizing it, this might unintentionally violate copyright laws and academic integrity.
Teachers need to help students understand that AI is not created out of thin air, but is a system built upon of human knowledge and culture.