Why Data and Software Should Be Protected like Candy

As machine learning and generative AI reshape the global economy, the legal frameworks governing their "ingredients", data and software, remain rooted in 18th and 19th-century paradigms. Current U.S. copyright and patent laws struggle to protect AI-assisted innovations, creating a "legal vacuum" that threatens investment.

This presentation uses a relatable "candy store" analogy to dismantle the complexities of AI intellectual property. We will explore why the final output (the candy), the platform (the storefront), and the training data (the ingredients) deserve dedicated legal protections.

A central focus of this presentation will be the Dual Protection Gaps:

1. The Copyright Gap: The "Authorship" requirement that excludes non-human expression.
2. The Patent Gap: The "Inventorship" requirement that denies protection to inventions where AI performs the "cognitive work" (Thaler v. Vidal).

The presentation proposes a middle-ground solution, introducing two new categories of intellectual property:

1. Data Rights
2. Software Rights

This framework is designed to reward commercial investment and technical architecture even when traditional human-centric standards are not satisfied.

Learning Objectives:

1. Identify the Authorship Gap: Explain why the U.S. Copyright Office refuses to register works without human authorship and how this affects AI-generated "Candy."
2. Analyze the Inventorship Gap: Deconstruct the Thaler v. Vidal decision and subsequent USPTO guidance requiring a "natural person" to provide a significant contribution to every claim in a patent
3. Apply the "Candy Store" Analogy: Use the physical world concept of "petty larceny" to explain why the unauthorized use of curated datasets and model architectures should be treated as a proprietary taking
4. Contrast Global IP Standards: Compare the U.S. "Human-Only" approach with international trends, such as the EU Database Directive, to assess risks of regulatory arbitrage
5. Evaluate Proposed Legislation: Critically examine the framework for 15-year "Data Rights" and "Software Rights" modeled after the Semiconductor Chip Protection Act