Washington's New AI Brain Trust: Inside the Private Group Shaping US Security

WASHINGTON, D.C. – July 09, 2026 – In a move that signals a new chapter in the fusion of technology and national security, a newly formed think tank, the Frontier Security Institute (FSI), has unveiled its inaugural class of fellows. The cohort is a veritable who's who of the American defense, intelligence, and technology establishment: 14 experts tasked with guiding the U.S. government through the treacherous landscape of advanced artificial intelligence.

This isn't just another advisory panel. The group includes a former CIA analyst who spent a decade tracking China's semiconductor ambitions, an award-winning wargame designer who has run crisis simulations for the White House, and a three-time founder specializing in securing software supply chains for defense systems. Their mandate is to help Washington protect its most powerful AI from theft, deploy it with confidence, and maintain an edge over global adversaries. The question is not just whether they can succeed, but how a private group operating behind closed doors will shape public policy on the most transformative technology of our time.

The New AI Praetorian Guard

The collection of talent FSI has assembled is formidable and speaks directly to the multifaceted nature of the AI threat matrix. The fellowship roster reads like the credits for a geopolitical thriller, featuring names like Gabriella Carney, a former senior policy advisor who architected U.S. export controls on semiconductors, and Dr. Todd Helmus, a former RAND scientist whose work on influence operations earned him a sanction from the Russian Federation.

They are joined by counterterrorism expert Dr. Sara Harmouch, DARPA veterans JC Herz and John M. Scott III, and the former Chief Scientist of the U.S. Army Research Laboratory, Dr. Joseph Mait. The group's collective experience spans the entire national security spectrum, from high-level policy and intelligence analysis to the granular details of systems engineering and warfighting assessments.

FSI's executive director, Isaac "Ike" Harris, emphasized the practical value of this expertise. "Frontier AI gives national security an extraordinary advantage, and these are the people who can help government seize it," Harris said in the announcement. "They've sat on the other side of the table — in program offices, on committees, in the field. They know how to turn that advantage into real decisions." The fellows will not be writing academic papers destined for dusty shelves; their role is advisory, channeling their deep operational knowledge into assessments, briefings, and, perhaps most critically, wargames for congressional staff, the intelligence community, and Pentagon program offices.

Bridging the Beltway's AI Gap

FSI's emergence addresses a critical vulnerability for the United States: the widening chasm between the hyper-speed of AI development in the private sector and the government's slower, more deliberative process of adoption and regulation. The term "frontier AI"—referring to the most powerful, next-generation models—represents a dual-use technology of unprecedented scale. These systems hold the promise of revolutionizing everything from military logistics to intelligence gathering, but they also introduce risks like autonomous cyberattacks, engineered pandemics, and sophisticated information warfare.

Washington is acutely aware of the stakes. Recent executive orders have focused on evaluating the offensive capabilities of new models, and a fierce debate is underway about whether to impose licensing regimes similar to those in the nuclear energy sector. FSI positions itself as a vital "translation layer" in this complex environment, helping national security operators understand, acquire, and govern AI systems being built in Silicon Valley and beyond.

To achieve this, the institute plans to operate through a combination of "third-party assessments, closed-door analysis and wargames, and neutral-ground convenings." The emphasis on closed-door sessions is key. It allows for candid discussions about classified threats and sensitive vulnerabilities, a necessity when dealing with national security. Wargaming, a specialty of several new fellows, provides a powerful tool to simulate how adversaries might exploit AI and to test U.S. countermeasures in a controlled, low-risk environment.

The Architecture of Influence

While FSI presents itself as an independent body, its structure and funding reveal a deeper connection to a specific corner of the AI world. The institute is housed within the Center for AI Safety (CAIS), a San Francisco-based non-profit founded in 2022 with a mission to reduce "societal-scale risks from artificial intelligence." This affiliation firmly places FSI within the "AI safety" movement, which often emphasizes worst-case scenarios and advocates for stringent guardrails on development.

CAIS is a 501(c)(3) organization, but it also operates a 501(c)(4) advocacy arm, the Center for AI Safety Action Fund. This dual structure is common in Washington, allowing the non-profit research side to feed analysis to a sister organization that can directly lobby policymakers. It suggests a clear pipeline from FSI's analysis to direct political influence.

Funding for this ecosystem comes not from the government, but from major philanthropic organizations. Public records show CAIS received a $4 million grant from Open Philanthropy and over $1 million from the Survival and Flourishing Fund. With reported revenues exceeding $10 million in 2024, the organization is well-resourced. This reliance on private philanthropy, while ensuring independence from government, raises its own questions about influence. The priorities of its donors, particularly those focused on existential risk, will inevitably shape the lens through which FSI views the intersection of AI and national security.

A New Model for Policy or a Black Box?

The rise of FSI is emblematic of a broader trend: the increasing reliance on private expertise to navigate complex, fast-moving technological challenges. In a world where a handful of tech companies control the development of frontier AI, the government arguably needs agile, expert partners to keep pace. FSI's model, bringing together seasoned veterans for targeted, confidential advice, could prove far more effective than traditional, slow-moving government commissions.

However, this model creates an inherent tension between security and transparency. The "closed-door" nature of FSI's most important work, while necessary for discussing sensitive intelligence, creates a black box around the formation of critical national policy. As this private brain trust advises on rules that will govern a technology impacting every citizen, the mechanisms for public oversight and accountability remain unclear. This mirrors concerns raised about similar bodies, like the UK's AI Security Institute, regarding the potential for industry capture and a lack of public scrutiny.

Ultimately, the Frontier Security Institute is a high-stakes experiment. It aims to provide the U.S. with a decisive edge in a global technological arms race by creating a rapid-response team of its best and brightest minds. The expertise of its new fellows is undeniable, but as their advice begins to filter through the halls of the Pentagon, Congress, and the intelligence agencies, their impact will be measured not only by the strategic advantages gained but by the transparency they afford the public they are tasked to protect.