WASHINGTON (AP) — New guidelines issued by the White House regarding the use of artificial intelligence (AI) by U. S. national security and intelligence agencies strive to balance the tremendous potential of this technology with the necessity of safeguarding against its inherent risks. The framework, signed by President Joe Biden and revealed on Thursday, aims to ensure that national security organizations have access to cutting-edge AI while also minimizing the chances of its misuse. Recent breakthroughs in AI are considered potentially transformative across various industries, particularly in military, national security, and intelligence operations. However, the use of this technology by the government poses risks, including the potential for mass surveillance, cyberattacks, or even the deployment of lethal autonomous systems. “This marks our nation's inaugural strategy for harnessing AI's power and managing its risks to enhance our national security, ” national security adviser Jake Sullivan explained while presenting the new policy to students at the National Defense University in Washington. The framework instructs national security agencies to broaden their adoption of advanced AI systems while explicitly banning certain applications, such as those that would infringe upon constitutionally protected civil rights or any system that would automate the use of nuclear weapons. Additional provisions promote AI research and emphasize the improvement of the security of the nation’s computer chip supply chain. The directives also urge intelligence agencies to focus on protecting the American industry from foreign espionage initiatives. Civil rights organizations have been monitoring the government's increasing reliance on AI and have voiced concerns regarding the potential for the technology to be misused. The American Civil Liberties Union (ACLU) stated on Thursday that the government is granting excessive discretion to national security agencies, enabling them to “police themselves. ” “While acknowledging the significant risks associated with AI, this policy falls short of adequately safeguarding us from perilous and unaccountable AI systems, ” stated Patrick Toomey, deputy director of the ACLU's National Security Project. “If the development of national security AI systems is an urgent priority for our nation, then establishing essential rights and privacy protections should be equally urgent. ” These guidelines were developed following a far-reaching executive order signed by Biden last year, which mandated federal agencies to formulate policies governing the use of AI. Officials emphasized that the rules are essential not only for promoting responsible AI usage but also for fostering the creation of new AI technologies, ensuring the U. S.

remains competitive with China and other nations also striving to leverage AI capabilities. Sullivan remarked that AI distinguishes itself from previous innovations—such as space exploration, the internet, and nuclear technology—that were primarily developed by the government. Instead, the advancement of AI systems has largely been driven by the private sector. He noted that AI is now “set to transform our national security landscape. ” Several figures from the AI industry who spoke to The Associated Press commended the new policy, describing it as a crucial step toward maintaining America’s competitive advantage. Chris Hatter, chief information security officer at Qwiet. ai, a tech firm that utilizes AI to identify vulnerabilities in computer code, expressed that the policy should garner bipartisan support. Without established guidelines, the U. S. risks falling behind in what he termed the “most consequential technology shift of our time. ” “The potential is immense, ” Hatter stated. “In military operations, we will witness autonomous weaponry—such as AI-driven F-16s and drones—alongside decision support systems enhancing human intelligence. ” Sullivan noted that AI is already altering how national security agencies handle logistics and planning, bolster cyber defenses, and analyze intelligence, suggesting that further applications may arise as the technology continues to evolve.