DARPA transitions innovative AI defense technologies from its GARD program to the Chief Digital and Artificial Intelligence Office, fortifying against adversarial attacks on AI systems.
Main Points:
- DARPA’s GARD program focuses on developing defenses against adversarial attacks, aiming to protect AI systems from being deceived or compromised.
- The initiative involves collaboration with industry partners and Google researchers, leading to the creation of new algorithms and open-source tools for broader community defense.
- Plans include transitioning GARD’s advanced defensive capabilities to other Defense Department components and investing in AI research for national security purposes, with a significant budget proposed for fiscal 2025.
Summary:
The Defense Advanced Research Projects Agency (DARPA) has taken a significant step in bolstering artificial intelligence (AI) defenses by transitioning cutting-edge capabilities developed under its Guaranteeing AI Robustness Against Deception (GARD) program to the Chief Digital and Artificial Intelligence Office (CDAO). The GARD program, initiated to counteract adversarial attacks on AI systems, has made strides in identifying and mitigating unique vulnerabilities inherent in AI technologies. Through collaboration with industry partners and Google researchers, DARPA has generated new algorithms and tools, now available as open-source for the wider community, thereby enhancing defenses across the AI and machine learning spectrum.
The transition to the CDAO is part of DARPA’s broader strategy to ensure these advancements benefit the entire Department of Defense (DoD), fostering an environment where AI and machine learning technologies are resilient against adversarial threats. The move aligns with DARPA’s mission to anticipate and mitigate strategic surprises through innovative research. Additionally, DARPA’s investment in AI, through initiatives like AI Next and the AI Forward initiative, underscores the agency’s commitment to advancing AI technology for national security. With a proposed budget of $310 million for fiscal 2025, DARPA aims to develop trustworthy, reliable, and ethically grounded AI technologies that meet critical national security and societal needs.
This transition and investment in defensive AI capabilities mark a significant milestone in enhancing the security and reliability of AI systems in adversarial environments, reflecting DARPA’s ongoing efforts to lead in AI innovation and application within the national defense arena.
Source: DARPA transitions new technology to shield military AI systems from trickery
Keep up to date on the latest AI news and tools by subscribing to our weekly newsletter, or following up on Twitter and Facebook.
