This week on IPWatchdog Unleashed, we tackle the AI arms race and how, if America really wants to lead in AI and emerging technologies, intellectual property policy cannot remain an afterthought.

My conversation is with Rama Elluru, who is Senior Advisor for the Special Competitive Studies Project, which is a bipartisan non-profit initiative, which makes recommendations to strengthen America’s long-term competitiveness for a future where AI and other emerging technologies reshape our national security, economy, and society. Elluru brings a rare cross-disciplinary perspective to the conversation, having worked as a computer scientist, patent attorney, Administrative Patent Judge at the U.S. Patent and Trademark Office (USPTO), and national security policy advisor. The discussion begins with her unconventional path into intellectual property, including her early work on embedded software tools for F-16 fighter jets, her clerkships at the International Trade Commission (ITC) and Federal Circuit, her private practice experience, and her work at the USPTO as AI began to emerge as a strategic policy issue.

After discussing Elluru’s background, our conversation then turns to the accelerating intersection of AI, patent law, and national competitiveness. We discuss whether the current U.S. patent system adequately incentivizes AI-related innovation, particularly as generative AI evolves toward more autonomous, agentic capabilities.

AI is no longer merely a commercial technology. It is not simply another software tool, another productivity platform, or another wave of enterprise automation. AI is rapidly becoming a foundation for economic power, military readiness, scientific discovery, cyber capability and geopolitical competition. The countries that get AI policy right will shape the next era of global influence. The countries that get it wrong will find themselves dependent on technologies, platforms and standards controlled by others.

In 2018, many people still viewed AI primarily through the lens of automation or software. Today, that framing is obsolete. AI is becoming a general-purpose technology with implications across drug discovery, cyber operations, defense systems, education, workforce development, infrastructure, communications and national security. The policy framework has not kept pace.

One of the most important parts of our conversation involved whether the patent system is properly positioned to incentivize AI innovation. Elluru was careful not to overstate the problem. With current AI tools, she agrees that the USPTO’s present inventorship framework is workable. When AI is being used as a tool or collaborator, the relevant inquiry remains whether a human made a significant contribution to conception. She said, “I do think that’s the right analysis right now.”

But the harder question is what comes next. Generative AI is already changing how researchers brainstorm, model, draft, test and iterate. Agentic AI—systems with more autonomous capabilities—could move the issue from speculative to urgent. Elluru raised the example of scientific discovery, including drug discovery and healthcare. What happens if a human identifies a disease target and an AI system autonomously searches a portfolio of molecules, runs simulations, tests possible solutions and produces a viable therapeutic candidate? Is the human’s identification of the problem enough? Did the AI contribute to conception? Can the result be patented? Should it be?

As I pointed out during our conversation, if the law remains locked into a framework that cannot accommodate increasingly autonomous innovation, we risk creating a perverse result: major breakthroughs may become difficult or impossible to protect. And if innovators cannot own the output, they won’t invest in developing it. I said, “If you can’t own it, you’re not going to be able to invest in it.” That is not merely a patent-law concern. It is an innovation-policy concern, a capital-formation concern and a national-competitiveness concern.

The United States cannot afford to sleepwalk into a future where its legal framework disincentivizes the very AI-enabled breakthroughs that will define economic and strategic leadership. Other countries will not wait. Elluru specifically noted that adversarial nations are watching these developments closely and may adjust their IP systems to ensure that AI-enabled innovation remains protected and incentivized. China, in particular, is not shy about treating intellectual property as a strategic tool.

That brings us to the larger issue: the persistent failure to connect intellectual property policy with national security.

I said during the conversation that it surprises me how many people still do not see the inextricable link between IP and national security. Elluru agreed, and her explanation was both practical and sobering. In her national security work she has encountered industry professionals who did not immediately understand why IP belonged in the conversation. They understood technology, defense and strategic competition, but patents and innovation incentives were not necessarily part of their operating framework.

That changed when IP leaders such as former USPTO Director Andrei Iancu and former Federal Circuit Chief Judge Paul Michel helped explain how intellectual property has shaped the American economy, from historic technology battles to biotechnology competition. Elluru explained that “IP is so important to our innovation and how innovation is very important to economy and how economy ties in with national security.” That is the chain policymakers need to internalize.

The United States often discusses IP theft as a national security problem, particularly when the issue involves China. But theft is only one part of the equation. A serious national strategy must also consider patent eligibility, inventorship, standard setting, trade secrets, enforcement, investment incentives and whether innovators can obtain reliable rights in frontier technologies. If the patent system becomes too uncertain, too expensive, too slow or too hostile to key categories of innovation, companies will respond rationally. They will increasingly rely on trade secrets. They may move activity offshore, under-invest, or decide that the American system is no longer the best platform for high-risk technological development.

That is why patent eligibility remains so consequential. Elluru identified confusing eligibility law as one of the issues she would prioritize if given the chance. She said, “our patent eligibility laws are still very confusing.” That confusion is not an academic inconvenience. In AI, diagnostics, software, biotech and data-driven innovation, eligibility uncertainty can distort investment decisions and weaken the U.S. position in strategically important sectors.

Our conversation also covered the risks of AI-enabled harm. Fraud, impersonation, cyberattacks and scams are likely to become more sophisticated as AI tools improve. Elluru mentioned using a family code word to guard against voice-cloning scams. I explained that I refuse to click payment links sent by text message even when they may be legitimate, because the fraud environment has made trust too expensive—and I hope that level of distrust does not infect the perception of AI. Sadly, when bad actors exploit technology, they degrade confidence in useful systems and slow adoption.

AI will create enormous benefits, but those benefits will not materialize automatically. Guardrails, enforcement and penalties for malicious use matter. So does public awareness. Elluru emphasized that people need to understand AI through direct experience rather than hype or headlines. A society that does not understand a technology will either overreact to it, under-regulate it, or regulate the wrong things.

The workforce challenge is equally real. Elluru acknowledged that AI may ultimately create new opportunities, as prior technologies always have, but she believes there will be near-term pain. Entry-level jobs may be disrupted. Education systems will need to adapt. Employers, universities and policymakers will need tighter feedback loops so students will graduate with skills that match the AI economy. The correct response is not to pretend disruption will not happen. The correct response is to build the transition architecture to help people move forward into the new AI economy.

Ultimately, this episode was about strategic seriousness. If America wants to lead in AI, it cannot treat intellectual property as a boutique legal issue. IP is not merely a private right. It is part of the national innovation stack. It determines whether capital flows into risky technologies, whether inventors disclose or conceal, whether startups can compete with incumbents, whether American companies can defend their breakthroughs, and whether the United States remains the preferred jurisdiction for building the future.

Elluru put the point plainly: “IP policy should be as strategic in our national frameworks” as foreign policy and defense policy. That should be the baseline for the next phase of the conversation because for too long, patent policy has been treated as a narrow matter for specialists. In an AI-driven world, the patent system is not peripheral infrastructure—it is a complete necessity. If the United States wants to win the future of AI, biotechnology, advanced computing, defense technology and scientific discovery, it needs an IP system built for that mission.

More IPWatchdog Unleashed

You can listen to the entire podcast episode by downloading it wherever you normally access podcasts or by visiting IPWatchdog Unleashed on Buzzsprout. You can also listen to IPWatchdog Unleashed conversations on the IPWatchdog YouTube channel. For more IPWatchdog Unleashed, see below for our growing archive of previous episodes.