Crafting effective AI policies for national security with insights from Anja Manuel

AI regulation for national security is especially nuanced and complex

May 15, 2024

Anja Manuel on AI national security and the need for global cooperation
The latest updates on AI legislation
Upcoming events, including Microsoft Build and DataConnect West
Before you go…check out the latest news about AI regulation

“It isn’t going to be a one-size-fits-all approach, and maybe some of the laws we already have in place should be specifically made to apply to AI.” - Anja Manuel

AI regulation is never simple, but in the realm of national security, it’s even more layered, nuanced, and complex. The possibilities for using AI to change the game (for the better) in national security are enticing, but it brings challenges that demand swift and strategic action.

In the latest episode of the RegulatingAI podcast, we discuss the pressing need for legislation that’s as dynamic as AI itself and examine how international cooperation is key to developing regulations that protect our global community while encouraging innovation.

We welcome Anja Manuel, the Executive Director of the Aspen Strategy Group and the Aspen Security Forum and Co-Founder and Partner at Rice, Hadley, Gates & Manuel, LLC.

Anja’s insights make the path forward clearer, framing effective AI legislation and emphasizing the need for global cooperation and ethical considerations. Her perspective—deeply rooted in national security expertise—underscores the critical balance between innovation and safeguarding against misuse.

Key takeaways:

The biggest potential risk areas for unregulated AI in national security are really all about how AI intersects with other technologies in possibly dangerous ways. The areas are: cybersecurity, autonomous weapons, bio (i.e., AI can come up with extremely toxic substances that don’t currently exist), and the control of nuclear weapons.
The best way to ensure legislation remains agile and responsive to some of these national security challenges is to never get behind in the first place—unlike how legislation and regulation lagged behind the rapid development of the internet and social media. Meanwhile, it’s important to be friendly to innovation. Fortunately, Manuel has observed, governments are mostly taking a thoughtful, moderate approach, and many legislators are making the effort to educate themselves. President Biden’s executive order from 2023 and the EU AI Act, are great starts in Manuel’s opinion.
There are opportunities for global cooperation on regulation. Existing political alliances and roadblocks will persist here. And autocracies will always be harder to work with. But there’s been a lot of cooperation—for example, the Chinese went to the UK AI Safety Summit, and the G7 put out the Hiroshima Process for code of conduct for advanced AI systems. Everyone is trying to figure it out.
Another reason for some optimism: Manuel believes National security agencies are “hugely” receptive to regulatory frameworks around AI. “What nobody wants here is an arms race that gets totally out of control,” she said.
She also sees public-private collaboration on national security considerations relating to AI happening. In contrast to a few years ago when Silicon Valley firms had no interest in working with the U.S. government, she says that younger tech companies, including AI companies, are thinking more carefully about how their AI can be used for good or not.
Of all the potential future scenarios or advancements in AI technology that could impact national security legislation, the one Manuel says she worries about the most is cybersecurity. That’s because it’s so relatively easy to create a sophisticated attack that causes huge damage—to, for example, the nation's infrastructure. It’s also commensurately easy to step up cyber defenses, but the scope of a successful attack is frighteningly large—much more than most any kinetic attack.
Manuel’s parting message for members of Congress, their staff, and think tanks: “I know international negotiations are slow and painful, and they often don't feel like they're getting somewhere. Keep. Talking. […] Continuing to talk to our international allies, but also our adversaries, is absolutely critical.”
Hear the entirety of Anya Manuel’s conversation with Sanjay Puri above and on Apple Podcasts and Spotify. Subscribe to the RegulatingAI podcast to never miss an episode.

The latest AI legislation

SB 21-190, Colorado, USA. Provides consumers the right to opt-out of the processing of their. Status: Enacted
[No bill number], New South Wales, Australia. Estalished in 2021 as "framework for capitalizing on critical technologies to drive a technologically-advanced, future-ready nation." Status: Discussion
5051, Brasil. The bill replaces three bills, Bill of Law 5.051/2019, Bill of Law 21/2020, and Bill of Law 872/2021, which were pending before the legislature over the past four years. Status: Discussion

Check out our live AI legislation tracker

Upcoming AI events

AI Expo 24 for National Competitiveness

Washington, DC | May 7-May 8, 2024 | In-person

SCSP, alongside the second Ash Carter Exchange on Innovation And National Security, is hosting this event at the Walter E. Washington Convention Center.

Designed to be the place to convene and build relationships around AI, technology, and U.S. and allied competitiveness, the AI Expo is meant to serve as a forum for industry, government, and academic research entities to exhibit some of the latest technological breakthroughs — in AI, biotech, energy, networks, compute, microelectronics, manufacturing, augmented reality, and beyond — and discuss their implications for U.S. and allied competitiveness.

Microsoft Build

Seattle, WA | May 21-23, 2024 | In-person, online

How will AI shape your future? Microsoft hosts its annual Microsoft Build conference at the Seattle Convention Center. Join Microsoft Build to grow your skills in topics like building copilots, generative AI, securing applications, cloud platforms, low-code, and more to unleash your creativity with the power of AI.

2nd Annual DataConnect West

Portland, OR | May 8, 2024 | In-person, online

Hear from industry experts with presentations and panels on AI/ML solutions that cover a variety of topics, including industry, ethics, education, sustainability, healthcare, society, and more. Stop by the Expo Hall during networking hour to learn how companies are utilizing and innovating with analytics. And join the social hour to conclude the day. Located at the Hyatt Regency Portland.

Bookmark this page to keep track of key upcoming AI events

Before you go…

EU and the Council Of Europe give the go-ahead for European AI Regulation | Mondaq
South Africa’s draft AI strategy needs more work, say pundits | ITWeb
US states lead the charge on AI regulation amid industry pushback | Fintech Global
First major attempts to regulate AI face headwinds from all sides | Times-Tribune

Bookmark this page to stay up to date on the latest AI regulation and policy news

Join our vibrant community!

Connect with experts, mentors, investors, and organizations to foster collaboration, innovation, and the responsible development of AI!