The development and training of AI models often necessitate the use of extensive datasets. When such data includes personal information, questions about permissibility and the ethics of its use arise. In addition, the use of both personal data as well as other categories of information (e.g. anonymized RWD) may be subject to certain restrictions, such as data protection requirements, intellectual property rights, or contractual restrictions (e.g. license agreements/T&Cs). It is crucial to navigate these challenges to ensure full compliance in AI development.

 In this interactive session, participants will be encouraged to explore and discuss considerations and possible adherence strategies that might be required for the lawful and ethical employment of available data for creating AI algorithms.

 Key Topics to be Addressed:

• Exploring the legality of utilizing data in AI algorithm training: Is it possible to repurpose existing datasets, such as those from clinical trials or RWD, for training AI algorithms, and how do we confirm compliance with relevant privacy legislation?
• Is the intended functionality and objectives of AI deployment a valid rationale for employing data in development processes?
• How may anonymization or pseudonymization be leveraged and what are its restrictions?
• What needs to be considered when scraping data from public sources