Copilot and training-data consent

There’s a question I’m very curious about: on one hand a tool like Copilot with genuinely excellent output, on the other the unauthorized use of the data these tools are trained on.

This dilemma looks set to be one of the most debated topics in generative AI over the coming years. The tool’s usefulness is undeniable, but the question remains: “what data was this output built on, and did its owner consent?”