Artificial intelligence continues to be a significant and disruptive technology. Investments in, and applications of, this technology continue to grow rapidly. The sophistication of large language models (LLMs) and their applications in critical processes, coupled with lingering concerns about their accuracy and tendency to “hallucinate,” have generated significant public debate about potential risks.

In science, AI has played a role in several important discoveries, including more-accurate-than-ever-before predictions of the structure and interactions of proteins, nucleic acids, small molecules, ions, and modified residues that have biological significance. Some researchers are concerned that AI will be employed in the design of unique new pathogens.

For defense applications, AI is increasingly applied to command and control, operational planning, logistics, autonomous systems, cybersecurity, and digital forensics, as well as modernizing core business operations. While logistics and planning applications are relatively benign, command and control applications may be problematic, especially in decisions to employ weapons, and especially in nuclear command and control. The head of the US Strategic Command recently stated that, while a human will always make the final decision on the use of nuclear weapons, it is conceivable that AI will be embedded in decision-support systems used for nuclear weapons. Even if a human is always in the decision loop, a too-heavy dependence on black-box systems could present a serious danger.

At the same time it is rushing to apply AI technology, the current US administration revoked a previous executive order on AI safety, reflecting a prioritization of AI innovation over safety and risk management that is mirrored in other major powers. Over the last year, the European Union’s AI act came into force, and the impact of that act is beginning to unfold, though the EU is under pressure from the US government and AI industry lobbyists to delay or roll back some of the act’s regulations.

Increasing chaos, disorder, and dysfunction in the world’s information ecosystem threaten society’s capacity to address difficult challenges, and it is clear that AI has great potential to accelerate these processes of information corruption. AI-enabled distortion of the information environment will likely remain an important obstacle to effective efforts to deal with urgent major threats like nuclear war, pandemics, and climate change. Large language model technologies and dramatic improvements in the phony video depictions known as deepfakes will have consequential future effects on the information ecosystem unless controls are introduced. Appropriate governance of AI and social media platforms is essential to an information ecosystem that supports truth and democracy; however, many media platforms are pulling back on commitments to moderate content and are reluctant to challenge incumbent political actors for fear of retaliation.

Indeed, in the United States formerly de-platformed purveyors of disinformation are now in positions of political authority and feel no qualms about exercising the powers of their new positions to push discredited narratives and pursue political enemies. The United States now has a president who personally participates in distributing fake information, most recently distributing AI-generated videos announcing a new health care concept—the medbed—that is a conspiracy theory based on false beliefs about UFOs. Both Russia and China have used such “deepfakes” and social media in their own wide-ranging disinformation campaigns.

In the United States, the scientific base out of which disruptive technologies emerge is under unprecedented attack, with arbitrary, ignorance-driven, meat-ax budget cuts and rescissions in research funding, attacks on the research infrastructure in universities, restrictions on public dissemination of vital data in areas relevant to climate change and public health, and prohibitions on government scientists publishing in the global scientific literature.

There is a growing belligerence among the United States, Russia, and China in space, and the probability of conflict in space continues to grow. China and Russia are far more active now than in previous decades, and US activities, both governmental and private, make it difficult to avert a military space race. The use of space systems—including privately owned Starlink satellites—to support military operations continues to expand. As a result, satellites—owned both by governments and corporations—become ever more important as military targets.

The Trump administration has announced plans for a nationwide “Golden Dome” defense against strategic ballistic missiles—essentially round two of the long-abandoned Reagan-era Strategic Defense Initiative, complete with space-based interceptors for boost-phase intercept of intercontinental ballistic missiles. Advocates argue that the technology context has changed dramatically since the 1980s, especially with respect to reduced space-launch costs and improvements in sensor technologies. However, as before, missile defense systems are at best only partially effective and serve as a provocation for the other side to invest in greater offensive capability. The danger is that Golden Dome will result in little real defense but will contribute to a deepening and dangerous arms race that extends to outer space. US Space Force leadership is now talking about space-based interceptors not only in terms of missile defense but also as elements of how the United States would conduct combat in space. At the same time, Russia and China appear to be contemplating the placement of nuclear weapons in space.

Read the 2026 Doomsday Clock statement »