Artificial Intelligence: A New Approach to Ethical Data Licensing

Artificial Intelligence and Data Licensing

The first wave of major artificial intelligence tools was trained on “publicly available” data, essentially scraping everything from the internet. Now, sources of training data are increasingly restricting access and pushing for licensing agreements. With the hunt for additional data sources intensifying, new licensing startups have emerged to keep the source material flowing.

Position of the Dataset Providers Alliance

The Dataset Providers Alliance (DPA), a trade group formed this summer, wants to make the AI industry more standardized and fair. To that end, it has just released a position paper outlining its stances on major AI-related issues. The alliance includes seven AI licensing companies, advocating for an opt-in system that requires explicit consent from creators and rights holders.

Shifting the Paradigm towards Ethics

• Alex Bestall, CEO of Rightsify, emphasizes that the opt-in rule is both pragmatic and moral.
• Industry leaders agree that current opt-out systems are unfair to creators.
• The DPA cautions against government-mandated licensing, promoting a free-market approach for negotiations.

Although the DPA’s ethical standards to source data could pose challenges, such as potential data scarcity, their guidelines suggest various compensation structures to incentivize creators. This could open avenues for fair payment models across different media forms.

Looking Ahead

As synthetic data becomes more prevalent in AI training, the DPA advocates for proper licensing of pre-training information to mitigate biases and ethical issues. The road forward may be complex, but the emergence of entities like the DPA indicates a significant shift toward ethical data usage in artificial intelligence.