Duke Officials Split on Speed vs. Safety in AI Push
Security leaders at Duke debated whether to accelerate AI development or impose strict federal regulations to prevent risks.
Key points
- NSA Deputy Director Tim Kosiba argues the US must prioritize speed in AI development to maintain national security.
- Rep. Deborah Ross advocates for federal regulations, including kill switch bills, to ensure AI safety and oversight.
- General Charles Brown warns that critical infrastructure is vulnerable to borderless cyber threats enabled by AI.
At the 2026 Cybersecurity at Duke conference, senior national security officials and legislators clashed over the appropriate pace for artificial intelligence development. The debate centered on whether the United States should prioritize rapid technological advancement to maintain global leadership or implement stricter regulatory frameworks to mitigate emerging risks. This division highlights a growing tension within the government regarding how to manage the dual-use nature of autonomous systems.
The event, hosted by the Pratt School of Engineering, featured Tim Kosiba, Deputy Director of the National Security Agency, and retired General Charles Brown Jr., former chairman of the Joint Chiefs of Staff. Both emphasized the urgency of securing AI infrastructure against cyber threats. They argued that hesitation could cede strategic advantages to adversarial nations, framing the current moment as a critical window for securing American technological dominance.
Security Leaders Prioritize Rapid Deployment
Tim Kosiba described the current AI landscape as an "Oppenheimer moment," suggesting that the US must act decisively to harness the technology’s potential. He compared the situation to historical periods of cryptographic innovation, such as World War II and the Cold War, where speed was essential for survival. Kosiba stated that the administration views speed as the primary objective, rejecting calls to pause research in favor of immediate execution and integration into national defense strategies.
General Brown echoed this sentiment, warning that cyber threats are borderless and increasingly targeted at critical domestic infrastructure. He cited recent attacks on municipal water systems as evidence of the vulnerability of physical networks connected to digital systems. Brown argued that relying on hope for safety is insufficient and that the nation must prepare for adversarial use of AI. He stressed that companies need clear policy guidance, but the government must also ensure that its own networks are hardened against sophisticated attacks.
Legislators Advocate for Regulatory Guardrails
Representative Deborah Ross, a Democrat from North Carolina, offered a contrasting perspective by emphasizing the need for congressional oversight. She advocated for legislation that would require AI companies to maintain the ability to slow down or shut down their models, a concept often referred to as a kill switch. Ross criticized the current lack of federal action, arguing that the administration’s reluctance to engage in regulatory processes leaves a dangerous gap in public safety protections.
The representative also promoted a bill that would establish a program within the National Institute of Standards and Technology for voluntary post-incident reporting. This mechanism would allow companies to share data on AI failures, enabling better tracking and industry-wide learning. Ross expressed frustration with the current political climate, stating that Congress is not currently acting with the necessary urgency. She suggested that future legislative sessions might see more decisive action if political control shifts, but warned that delay could result in irreversible consequences.
Shared Need for Skilled Leadership
Despite their differing views on regulation and speed, all three speakers agreed on the importance of human expertise. They emphasized that technology alone does not create a sustainable advantage without educated and talented leaders who can navigate complex ethical and technical challenges. This consensus underscores the broader need for a workforce that understands both the capabilities and the risks of autonomous systems.
The discourse at the Duke conference reflects a broader national debate about the role of government in shaping the AI landscape. As the technology continues to evolve, the trade-off between competitive speed and regulatory safety remains unresolved. The outcome of this debate will likely determine the future trajectory of AI development in the United States and its impact on global security dynamics.
