The whitepaper explains the legal risks and responsibilities when using generative AI tools in software development. It highlights confidentiality concerns, noting that enterprise tiers and opt-out settings prevent training on client data. It addresses IP ownership, stating most tools assign rights to the developer, but unresolved issues—such as the GitHub Copilot lawsuit—create uncertainty about training data and attribution. Pages 14–16 compare open-source LLMs with hosted APIs, emphasizing transparency, control, and reduced vendor lock-in. The paper recommends strong governance: choosing tools carefully, ensuring data security, disclosing usage to clients, mitigating bias, and reviewing code for infringement.

Join for free to read