The OpenAI and Codex Lawsuit: Everything You Need to Know Explained By OpenAI Itself

MEMORANDUM OF POINTS AND AUTHORITIES

I. INTRODUCTION

This case is about an emerging field of artificial intelligence known as “Generative AI.” Generative AI systems learn concepts and relationships from large bodies of existing knowledge, and use what they learn to help people create new works.

Plaintiffs’ lawsuit challenges two generative AI tools that help people write computer programming code—GitHub Copilot and OpenAI Codex. But Plaintiffs attempt to plead causes of action that don’t actually apply to these tools, rendering the complaint subject to dismissal on multiple grounds.

OpenAI is the creator of Codex and a pioneer in the field of AI. It is an independent company whose mission is to ensure that artificial intelligence, including generative AI systems, benefit all of humanity. OpenAI is governed by a non-profit and its organizational structure limits the economic returns of investors and employees.

Open AI developed Codex using publicly available computer programming code based on theories of how the human brain learns from new information. With that training, Codex generates coding suggestions in response to a person’s requests. A programmer can provide a short text request (e.g., “create a button on a website that lets a user upload a document”), and Codex will generate a coding suggestion to meet the request.

Codex’s training allows it to help people write code for common functions and for functions that have never been written before. This capability makes computer programming more accessible to broad segments of the population and makes those who have already learned to program more productive.

The essence of Plaintiffs’ complaint is that rarely—the complaint cites a study reporting 1% of the time—Copilot (and therefore Codex) allegedly generates snippets of code similar to the publicly available code that it learned from, and does so without also generating copyright notices or open source license terms that originally accompanied the code.

But Plaintiffs provide no allegation that any code that they authored was used by Codex or generated as a suggestion to a Codex user; they only point to Codex’s abilities to generate common textbook programming functions, such as a function for determining if a number is odd or even. Plaintiffs also do not allege copyright infringement by any OpenAI Entity, but instead allege a grab bag of claims that fail to plead violations of cognizable legal rights.

The complaint thus suffers from both threshold defects requiring dismissal of the entire complaint and specific defects in each pleaded cause of action.