Foreign influence is a persistent threat to U.S. elections, and three countries — China, Iran, and Russia — are poised to attempt to influence the 2026 midterms. Though their behavior is not new, this year, their access to sophisticated artificial intelligence tools and the lack of federal governmental pushback against them will likely require additional diligence from voters to ensure that they have accurate information.
The Trump administration has degraded the federal infrastructure for detecting and countering foreign election threats, leaving states to prepare for these threats largely on their own for the first time since 2016. The rapid advances in artificial intelligence tools have increased both the capabilities of existing actors and made the influence campaigns more believable. Further, domestic actors, including the Trump administration, are using the real threat of foreign influence — mixed with promotion of unsubstantiated rumors of foreign interference, meaning the technical targeting of voting machines or other hardware — to advance their political goals.
Americans should know that they are being targeted by foreign propaganda and should rely on trusted sources, including local election officials and well-established news sites, to validate facts about the midterm election processes and administration. Voters should also be skeptical of novel claims by the Trump administration alleging foreign meddling in the 2020 or future elections to justify illegal federal control over elections.
For the first time since the designation of election infrastructure as critical infrastructure, the federal government’s support is notably absent this cycle, representing a new challenge for securing elections from foreign threats. Federal agencies are no longer funding security improvements, providing cybersecurity support for elections, or sharing threat information about foreign actors. (The Brennan Center is documenting the dismantling of federal support, which has undone years of threat detection and victim notifications regarding foreign influence or interference operations.)
This change is a significant vulnerability going into the 2026 midterms, leaving election officials in the dark about foreign actors’ plans and actions. Unlike official statements in prior election cycles, then-Director of National Intelligence Tulsi Gabbard did not mention foreign election threats in her worldwide threats testimony and accompanying unclassified report. Further, her office dragged its heels in making preparations that are usually in place at the beginning of an election year. It finally appointed officials to fulfill the role of the election threats executive in May. This decision comes after the director of the National Security Agency and Cyber Command acknowledged that it has not yet established the Election Security Group, a blow to threat detection and defensive cyber measures.
In previous years, the federal government mobilized to respond to foreign influence efforts transparently and in coordination with key actors, such as election officials, to mitigate the impact and build public trust in the security of U.S. election systems. Its withdrawal places a greater burden on local election officials and civil society organizations, which have previously relied on the unique capabilities and expertise of the federal government to detect and communicate about covert or clandestine foreign threats.
Even without the federal government’s up-to-date insights on the intentions and capabilities of China, Iran, Russia, and other nation-states, we know those countries have only continued to improve their tactics and techniques since the 2024 election. Technology companies, civil society, and nongovernmental organizations are observing active and ongoing global Chinese, Iranian, and Russian influence campaigns.
In March, the European Council on Foreign Relations reported that the Chinese government paid individuals to pose as Western journalists or influencers to parrot its propaganda, a tactic also used by Russia in 2024. Russia’s Social Design Agency, a state-controlled public relations firm, hacked users on Bluesky to promulgate fake news stories, an evolution of its prior technique of creating fake online personas to spread propaganda. In May, the Europe-based investigatory journalism organization Organized Crime and Corruption Reporting Project revealed the role of the Social Design Agency in organizing false flag vandalism in Europe in attempts to incite unrest, as well as in running information operations targeting Russian-speaking voters in Armenia’s parliamentary elections.
Iran’s propaganda machine has demonstrated growing sophistication with the outbreak of war at the end of February, producing videos that attempt to confuse the public about what is actually happening on the ground. There is little information indicating that its cyber capabilities are diminished from the war, and it’s likely that Iran will use all available tools to undermine the United States and trust in the midterm elections.
Another challenge this election cycle is the widespread availability and increasing capabilities of AI tools. These can generate more persuasive influence operations and reduce the barrier to entry for new actors that did not have the resources to target the United States in the past. Old and new conspiracy theories are likely to flourish in quantity and quality, making them appear more authentic to public audiences. Foreign actors regularly use AI tools such as chatbots, large language models, video-modification technology, and voice cloning in their influence campaigns and planning.
In February, OpenAI observed an individual linked to the Chinese Communist Party prompting ChatGPT to design a multistep influence operation using several tactics targeting Japanese Prime Minister Sanae Takaichi. ChatGPT refused to generate the plan or provide feedback on this operation for the prompter, but such barriers may not occur in every model or in every case of attempted influence.
Meta reported that with the help of AI, Iran is conducting more compelling campaigns using familiar influence tactics. Iran created fake news websites to amplify false content and used state media organizations or AI-created influencers to distribute false information. In February, Iran created a fake video of a U.S. fighter jet crash that was amplified by news outlets associated with the Chinese and Russian governments. While some of the tactics have remained the same compared with prior years, the volume and dissemination of content are potentially increasing because of AI. Actors also benefit from AI’s ability to rapidly process large amounts of data and information, performing analysis at speed and scale. The China-based Golaxy Labs, for instance, can be used to enhance the relatability of false messages for targeted audiences and allows actors to flood platforms with content.
Still, there is some good news. While influence operations are increasingly difficult to detect, election interference — targeting technical electoral systems — remains challenging for foreign governments and actors. This is because U.S. electoral systems are diverse, and security investments made from 2018 to 2024 have steadily fixed vulnerabilities where possible. Further, the systems that contain vote-casting and tabulation functions are often stored offline or on separate networks, reducing the chances that remote actors living abroad can hack the machines.
Nevertheless, in the current polarized political climate, rumors of election interference could be used or amplified by federal officials, as Trump has installed election deniers in high positions throughout the federal government. False claims or unfounded suspicions about foreign election interference have already been used to justify unprecedented meddling by federal agencies in elections, a problem that could increase in the run-up to and aftermath of Election Day.
• • •
