Artificial intelligence is moving from a tool used by people to an independent force in political communications, raising fresh concerns about misinformation and the integrity of election discourse as the U.S. prepares for midterm voting.
AI-generated images and autonomous propaganda
Election watchdogs and Democrats in New York have flagged the use of AI-generated images in campaign advertising, including ads that depict Gov. Kathy Hochul. Some of those ads were not labelled to indicate they were produced with artificial intelligence, prompting calls for legal scrutiny of how such material is presented to voters.
Beyond manipulated images, researchers say a new breed of automated accounts — sometimes described as “supercharged” bots — can now operate with minimal human direction. A study from the University of Southern California cited by experts describes systems that can set goals, assemble teams of automated agents, generate content and learn from their results, producing posts that vary slightly and therefore appear more authentic to the public.
“What we know is that it’s actually quite easy to generate this content and distribute it through social media platforms, so it’s a great concern,”
The comment came from Jennifer Stromer-Galley, a professor at Syracuse University’s School of Information Studies, who has been tracking the growing sophistication of these tools and their ability to amplify disinformation without continuous human oversight.
Why the new generation of bots matters
Analysts say the evolving capabilities of AI-driven systems change the dynamics of online persuasion and information spread in several ways:
- Scale: AI makes it cheap and fast to produce vast quantities of content.
- Variety: Algorithms can rephrase and reshape messages so they appear to come from diverse, independent sources.
- Autonomy: Advanced systems can set their own objectives and adapt strategies without direct human micro-management.
Those features mean bad actors can more easily manufacture the appearance of consensus or grassroots sentiment, complicating the public’s ability to distinguish genuine debate from coordinated influence operations.
| Reported capability | What it enables |
|---|---|
| Setting goals | Automated campaigns can prioritise objectives such as maximising shares or targeting specific demographics. |
| Building teams | Multiple agents can coordinate to amplify a message across accounts and platforms. |
| Writing posts | Systems generate original social posts, varying text to reduce detection and mimic human authorship. |
| Learning from mistakes | Algorithms adjust tactics when content underperforms or triggers moderation. |
Regulatory and practical gaps
Despite the growing role of AI in political messaging, there is no federal U.S. legislation in place that specifically regulates the use of AI in political advertising or requires standardised disclaimers. Efforts led largely by Democrats to introduce constraints — including bans or mandatory labelling for AI-produced political content — have not cleared Congress.
That regulatory vacuum leaves the responsibility for detecting and responding to deceptive AI-driven content split among social media platforms, state election bodies and political opponents. In New York, state Democrats are urging the state Board of Elections to pursue legal action over alleged unlabeled AI images used in campaign ads in Nassau County.
Experts warn that platforms and regulators must move faster to address the threat. Because research indicates the public often cannot reliably tell AI-generated content from human-created material, the potential for confusion and manipulation is significant.
Implications for democratic discourse
The combination of powerful generative tools and autonomous coordination changes the calculus for campaigns, regulators and the public. Where once disinformation operations required a human workforce to produce and distribute content at scale, they can now be compressed into software that designs, tweaks and amplifies narratives automatically.
That shift has three practical consequences: it increases the speed and volume of potentially misleading content, it makes coordinated campaigns harder to attribute, and it lowers the cost‑barrier for actors seeking to influence public opinion.
As democratic systems grapple with these risks, the absence of national rules in the U.S. underscores both the technical challenge and the political difficulty of balancing free expression, innovation and election integrity in the age of AI.
For policymakers and platforms, the debate now centres on whether disclosure rules, tighter ad labelling, platform policy changes or new legislation can meaningfully reduce the harms posed by autonomous AI-driven political communications.
