As AI-generated content becomes ever more common, AI’s limitations are becoming more apparent. Lawsuits are raining down like proverbial cats and dogs.

Well, here’s the good news if you’ve been caught with your fingers in the AI cookie jar - artificial intelligence (AI ) behemoths OpenAI have created their Copyright Shield, a program that promises legal refuge for users, while Anthropic, backed by tech behemoths, banks on a liability-shifting model. As the courtroom drama unfolds, who bears the brunt of AI-generated creativity?

Hitting the Courts

Why is all this relevant? Well, Anthropic is currently being sued for copying song lyrics, and industry leader OpenAI is being done for training its models on real life books.

OpenAI and Copyright Shield

Perhaps in a bid to stay one step ahead, OpenAI has played its wildcard in the form of Copyright Shield. During the company’s recent dev conference, CEO Sam Altman, promised to foot the legal bills for users accused of breaching copyright laws, if they’re paying customers and not just skimming off the free version of ChatGPT.

A little on OpenAI’s Copyright Shield:

Something I’ve learned over the years is that to big companies, skirting & breaking laws and being fined or sued after the fact is part of the cost of doing business. A thread: 1/🧵 https://t.co/Mwkn6c0fkG
— Jingna Zhang (@zemotion) November 9, 2023

"We can defend our customers and pay the costs incurred if you face legal claims around copyright," said OpenAI CEO Sam Altman in his keynote speech. That’s certainly a novel idea in a landscape where AI's creative feats dance on legal tightropes.

Anthropic's Liability Ballet: A Risky Dance

Anthropic and its tech titan backers, on the other hand, is doing something different. While they argue that copying text to create new material isn’t inherently bad, they’re also claiming that their systems were specifically designed to avoid copyright claims.

However, where claims are made, Anthropic Deputy General Counsel Janel Thamkul says, "Responsibility for a particular output will rest with the person who entered the prompt to generate it. That is, it is the user. If we detect repeat infringers or violators, we will take action against them, including by terminating their accounts."

Hmmm…

First things first, don’t infringe someone’s copyright and don’t steal their work. But if you’re going to, it seems best to use ChatGPT. We're off to lament the death of creativity.