Why Dorsey and Musk are targeting IP law now

[Updated with quotes from Matthew Asbell, an IP attorney and partner at Lippes Mathias.]

When tech heavyweights Jack Dorsey and Elon Musk recently advocated to “delete all IP law,” the exchange on X caused ripples beyond social media. The comment originated with Dorsey, Twitter’s co-founder, to which Musk—X’s owner—replied, “I agree.” This taps into long-standing skepticism toward intellectual property in some tech circles. The controversy highlighted a key tension: Can established IP systems adapt to the rapid pace of technological change, especially regarding AI training data? And should Dorsey’s and Musk’s calls be taken literally?

The calls could push for radical IP reform

There might be more to the story. Both Dorsey and Musk’s companies hold patents. Dorsey later clarified his views, criticizing existing IP frameworks as “rent-seeking” systems that unfairly benefit intermediaries over creators. He argued for “much greater models to pay creators,” suggesting a systemic overhaul rather than simple deletion.

This discussion occurred as legal pressure on AI developers intensified. High-profile lawsuits, like one from The New York Times, alleged copyright infringement in training data. Tech giants like Meta invoked fair use defenses. Shortly before Dorsey’s and Musk’s comments, twelve major copyright lawsuits against OpenAI and Microsoft were consolidated in New York. Musk is also separately suing OpenAI, arguing it deviated from its original non-profit mission.

‘Delete IP’ call prompts blowback

The “delete IP” call sparked polarized reactions. Supporters saw it as challenging outdated restrictions hindering innovation. Bitcoin advocate Stephan Livera framed IP laws as restrictive monopolies. Conversely, critics called it an attack on creators’ rights. Rumble CEO Chris Pavlovski contended that eliminating IP would let big tech exploit creators: “Creators’ IP must be protected… you just proposed to destroy that world,” he posted on X. Ed Newton-Rex, founder of an ethical AI training non-profit, described it as “Tech execs declaring all-out war on creators.” Others, like writer Lincoln Michel, pointed out the perceived hypocrisy of tech leaders who built value using IP now wanting to dismantle it. Attorney Nicole Shanahan argued IP law is essential, though open to reform.

Responding to common criticisms leveled against the legal profession during such debates, IP attorney Matthew Asbell, a partner at Lippes Mathias, pushes back: **“It’s overly simplistic to say IP attorneys oppose a ban only to save our own jobs. We fight for the principle that society prospers when creativity is rewarded and knowledge is shared under clear rules.”** Asbell also highlights the fundamental nature of such a change:

Abolishing IP isn’t a minor tweak—it would mean rewriting Article I, Section 8 of the Constitution and reinventing America’s role in the world economy.

Musk’s agreement aligns with his complex history regarding IP. He has voiced skepticism, calling patents “for the weak” and “landmines” hindering progress. SpaceX largely avoids patents, relying on rapid innovation and secrecy, partly fearing patents act as a “recipe book.”

Yet while Tesla pledged in 2014 not to sue those using its patents in good faith, the company maintains and expands a large patent portfolio (over 2,900 globally as of 2022), with only some covered by the pledge. Asbell points to this apparent contradiction: **“Tesla alone holds 90 U.S. trademarks and 579 patents, about 15 naming Musk as inventor—proof that the very entrepreneurs calling for abolition built their empires on IP protections.”** Earlier ventures like PayPal also relied on patents and trademarks. Musk’s endorsement likely indicates a desire for radical reform of systems he views as barriers, not literal abolition overnight.

This anti-IP stance isn’t universal among tech pioneers. Bill Gates advocated for software copyright early on. In his 1976 “Open Letter to Hobbyists,” he argued that without paid licenses, software innovation would stall. Decades later, Apple under Steve Jobs pursued an aggressive IP defense strategy, viewing patents as core assets and famously threatening legal action against competitors like Android.

AI exposes a widening gap between tech’s open-source ethos and IP-driven sectors

This historical tension is reflected in modern software development, especially with AI. GitHub—owned by Microsoft—hosts millions of repositories, mostly under permissive open-source licenses (like MIT). Yet, generative AI creates friction. Training models requires huge datasets, often scraping copyrighted or open-licensed works, sparking legal fights over fair use and derivative works—issues traditional licenses didn’t anticipate. This fuels conflict between proprietary AI (like early OpenAI) and open-source alternatives. Microsoft navigates this complexity: it manages GitHub and partners with OpenAI (though recent moves suggest a less exclusive relationship), while also exploring various open and closed models. This situation highlights the conflicts between established IP norms and the drive for rapid AI innovation surfaced by the Dorsey/Musk exchange.

The tech world’s often skeptical view of IP contrasts sharply with its role in other R&D-heavy sectors like pharmaceuticals. The pharma industry deems patent protection essential. Bringing a drug to market can cost billions and take more than a decade of research and trials. Without the market exclusivity patents offer, the financial incentive for these costly projects would diminish. Industry groups like PhRMA consistently advocate for strong global patent enforcement, a view shared by biotechnology and advanced chemicals, where large investments depend on IP for returns. For these sectors, IP is the engine for funding development.

Japan, the EU, and China test conflicting approaches to AI and intellectual property

Beyond U.S. legal battles, other economies are developing distinct AI and IP regulations. Japan was called a “machine learning paradise” owing to its Copyright Act’s Article 30-4, which broadly permits using copyrighted works for AI training without permission, unless the main purpose is direct “enjoyment” of the work. Yet this approach faces pushback. Some interpretations emphasize rights holder protections gaining ground, and potential reforms discussed in early 2025.

In contrast, the European Union’s AI Act takes a more cautious approach. While not classifying most generative AI as “high-risk,” the Act requires transparency. It notes that developers must publish summaries of copyrighted training data. Developers must also comply with the EU Copyright Directive, which allows rights holders to opt-out of data mining via machine-readable signals. This regulation thus prioritizes rights holder control. It may, however, limit data access and potentially hinder European AI competitiveness.

China approaches AI and IP with something of a bifurcated approach. On the one hand, it leads globally in filing generative AI patents (and patents overall) while also fostering homegrown large language models (like DeepSeek R1’s, prominent in early 2025) released under open-source licenses. This strategy uses patents traditionally while leveraging open source to speed up domestic AI adoption, possibly bypass hardware limits, and assert global tech leadership—a distinct use of both IP protection and open innovation for national goals.

These diverging international strategies—Japan’s debated permissiveness, the EU’s regulated caution, China’s strategic duality—show the global challenge of adapting IP for AI. As Matthew Asbell warns regarding the potential consequences for the U.S., **“Scrapping IP would cede leadership in science and tech to nations doubling‑down on patents; America would swap creativity for factory shifts—an economy from ‘the China of previous decades.’”**

Dorsey’s call for “much greater models to pay creators,” alongside his critique of current systems, mirrors a widespread search for sustainable models that do two things. One, meet AI’s data needs and second, foster innovation while simultaneously compensating the human creativity.

These national approaches are large-scale experiments testing different balances between encouraging technology and protecting rights. Progress under diverse systems, including China’s strategic open-source use or Japan’s rapid AI development, challenges the assumption that strong IP protection is the only way to drive technological progress. But given the long history of intellectual property protections, together with the rapid growth of patent filing in regions such as Southeast Asia, it may be premature to expect a collapse of IP systems anytime soon. Instead, we are likely witnessing the beginning of a global recalibration. Pushing the argument to its extreme, Asbell provocatively asks:

If they truly believe this, why stop at IP? Let’s abolish land ownership and go fully communist—a curious proposal from men who thrived in capitalism.