There is a beautiful irony unfolding in a California courtroom. Midjourney, sued by three Hollywood studios for training on copyrighted images, is now compelling those same studios to reveal exactly how much AI they use themselves. Discovery as judo. The studios built their case on the premise that AI companies exploit creative labor. Midjourney is betting the studios are doing the same thing, quietly, at scale.
The AI Transparency Double Standard
Meanwhile, Alibaba has classified Claude Code as high-risk software, banning internal use entirely. The reasoning reported is security and data leakage. But read the move alongside the Hollywood lawsuit and a clearer picture emerges: institutions that profit from AI are simultaneously restricting AI access for their own workers and obscuring their AI usage from the public. Fast Company's piece on why telling employees to adopt AI does not work adds another layer. Organizations are mandating AI use at the strategic level while creating compliance theater at the operational one. Nobody is being honest about what they are actually doing with these tools.
Disclosure as the New Copyright
What Midjourney's legal strategy actually invents is a disclosure framework. If Hollywood must reveal AI usage in its productions to defend its copyright claims, that precedent ripples outward into advertising, journalism, and music. The Fast Company piece on measuring AI's business impact notes that most enterprises still have no coherent metrics for what AI is doing inside their workflows. You cannot disclose what you do not measure. Beeple and Jehan Chu's conversation on digital art as the visual language of everything we see feels relevant here: when the pipeline is invisible, authorship becomes a legal fiction.