Computer scientists have long sought to build systems that can actively and autonomously carry out complicated goals in the real world—commonly referred to as artificial intelligence "agents." Recently, significant progress in large language models has fueled new optimism about the prospect of building sophisticated AI agents. This CSET-led workshop report synthesizes findings from a May 2024 workshop on this topic, including what constitutes an AI agent, how the technology is improving, what risks agents exacerbate, and intervention points that could help.

View this session of our Security and Emerging Technology Seminar Series on August 1 at 12 p.m. ET. This session featured a discussion on the President’s Council of Advisors on Science and Technology (PCAST) Report on Strategy for Cyber-Physical Resilience.

CSET's Josh A. Goldstein was recently quoted in a WIRED article about state-backed hacking groups using fake LinkedIn profiles to steal information from their targets. Goldstein provides insight by highlighting the issues in the disinformation space.

Militaries seek to harness artificial intelligence for decision advantage. Yet AI systems introduce a new source of uncertainty in the likelihood of technical failures. Such failures could interact with strategic and human factors in ways that lead to miscalculation and escalation in a crisis or conflict. Harnessing AI effectively requires managing these risk trade-offs by reducing the likelihood, and containing the consequences of, AI failures.

A report by CSET’s Josh Goldstein, Micah Musser, and CSET alumna Katerina Sedova in collaboration with OpenAI and Stanford Internet Observatory was cited in an article published on Medium. The report explores how language models could be misused for influence operations in the future, and it provides a framework for assessing potential mitigation strategies.

Funding and priorities for technology development today determine the terrain for digital battles tomorrow, and they provide the arsenals for both attackers and defenders. Unfortunately, researchers and strategists disagree on which technologies will ultimately be most beneficial and which cause more harm than good. This report provides three examples showing that, while the future of technology is impossible to predict with certainty, there is enough empirical data and mathematical theory to have these debates with more rigor.

CSET Senior Fellow Andrew Lohn testified before the House of Representatives Homeland Security Subcommittee on Cybersecurity, Infrastructure Protection, and Innovation at a hearing on "Securing the Future: Harnessing the Potential of Emerging Technologies While Mitigating Security Risks." Lohn discussed the application of AI systems in cybersecurity and AI’s vulnerabilities.

CSET Senior Fellow Andrew Lohn testified before the House of Representatives Science, Space and Technology Subcommittee on Investigations and Oversight and Subcommittee on Research and Technology at a hearing on "Securing the Digital Commons: Open-Source Software Cybersecurity." Lohn discussed how the United States can maximize sharing within the artificial intelligence community while reducing risks to the AI supply chain.

CSET Senior Fellow Andrew Lohn testified before the U.S. Senate Armed Services Subcommittee on Cybersecurity hearing on artificial intelligence applications to operations in cyberspace. Lohn discussed AI's capabilities and vulnerabilities in cyber defenses and offenses.

Based on an in-depth analysis of artificial intelligence and machine learning systems, the authors consider the future of applying such systems to cyber attacks, and what strategies attackers are likely or less likely to use. As nuanced, complex, and overhyped as machine learning is, they argue, it remains too important to ignore.