US Administration Toughens Stance on AI Copyright Protections
- •Defense Undersecretary Emil Michael accuses Anthropic of massive copyright and identity data theft.
- •White House reinforces 2025 Executive Order requiring AI models to respect existing copyright laws.
- •Shift in federal policy suggests waning influence of tech lobbying regarding AI training data.
A late-night social media post by Undersecretary of Defense Emil Michael has sparked debate regarding the relationship between the federal government and AI laboratories. In a now-deleted message, Michael accused Anthropic of failing to respect intellectual property, labeling the company a "thief" of both public identity data and creative works. This friction arises from a contract dispute where the military seeks access to public databases that the firm allegedly attempted to restrict.
The incident underscores a shift in the Trump Administration's approach to AI regulation. While Silicon Valley has lobbied to frame copyright flexibility as a national security imperative, the White House has reinforced its December 2025 Executive Order. This directive mandates that AI development must align with existing copyright laws, challenging "fair use" justifications frequently cited by developers who train models on scraped internet data (unauthorized collection of website information).
Michael’s history as a tech executive adds weight to these criticisms, reflecting an intimate understanding of Silicon Valley culture. His mention of ongoing lawsuits highlights the mounting legal pressure on AI firms, including potential liability for utilizing pirated datasets. This stance suggests an end to regulatory leniency, prioritizing creator rights over the expansion of synthetic media tools (content generated by artificial intelligence).