 the public’s imagination in 2023, as Americans experienced novel applications such as chatbots and image generators. AI will almost certainly continue to evolve at a rapid pace in the years to come, with public and private entities around the world vying for profit and competitive advantage. The evolving AI landscape will present cyber defenders with new opportunities to defend critical infrastructure against malicious activity. The cybersecurity community has a long history of harnessing the power of machine learning techniques for basic tasks like data processing, email filtering, and malware identification. New cyber defense tools that integrate AI could eventually enable cyber defenders to more efficiently detect anomalous network traffic and other adversary activity, coordinate the defense of complex systems and networks, and augment a cybersecurity workforce that is already stretched thin. AI tools may also make our software development ecosystem safer and more secure. While LLMs have shown some fluency in programming languages, they cannot yet generate commercially useful secure code without human intervention. Responsible integration of AI tools into the software development lifecycle may enable developers to identify vulnerabilities in new code and suggest potential fixes. As AI tools mature, they could be able to make widely used software products more secure by rewriting existing code into a memory-safe programming language. However, realizing the promise of AI also challenges us to manage the risks it poses to cybersecurity. Today, LLMs can quickly and cheaply generate persuasive and micro-targeted text, images, audio, and video in different languages. Cybercriminals, hacktivists, and others with limited resources and technical sophistication may use these capabilities to conduct phishing campaigns, information operations, and other malicious cyber activity. AI-enabled surveillance and censorship technologies enable authoritarian regimes to more effectively and efficiently target journalists, dissidents, and human rights defenders. Without safeguards, AI technologies may also put Americans’ privacy at risk by making it easier to extract, identify, and exploit personal data. As the AI ecosystem continues to evolve, there is an opportunity to ensure that its core elements—data, computing, and algorithms—are developed with safeguards against misuse. 2024 REPORT ON THE CYBERSECURITY OF THE UNITED STATES POSTURE 7