Overview of Bringing a chatbot to the polls
This NPR It's Been a Minute episode examines how AI chatbots and social media influencers are shaping election information and voter behavior as the 2026 midterms get underway. Host Brittany Luce speaks with NPR politics reporter Maham Javed and media scholar Dr. Martin Riedel about why voters are turning to chatbots like ChatGPT for ballot research, why campaigns are increasingly using influencers, and what the ethical and practical risks are when political information comes from systems designed to be persuasive, fast, and highly personalized.
Why voters are turning to AI chatbots
What people are using them for
Voters are using chatbots as a kind of personalized voting assistant to:
- compare candidates and ballot measures
- verify claims they see elsewhere
- get quick summaries of down-ballot races
- fill gaps left by shrinking local news coverage
Why chatbots feel useful
According to Javed, people like that chatbots can:
- put all candidate information in one place
- generate side-by-side comparisons
- feel immediate, convenient, and easy to use
- fit into habits they already have for daily life tasks like planning, email, and search
The episode notes that many users don’t see this as unusual—they already rely on AI for lots of everyday decisions, so using it for voting feels like a natural extension.
How influencers have become political actors
Why campaigns use influencers
Dr. Riedel explains that political campaigns increasingly lean on influencers because:
- young voters often get news through social feeds
- audiences trust creators with established relationships
- influencers can create the appearance of broad, organic support
- campaigns want to “meet voters where they are”
What makes influencers different from celebrities
The conversation distinguishes between traditional celebrity endorsements and modern influencer politics:
- celebrity endorsements have existed for decades
- micro-influencers can have smaller audiences but much higher engagement
- influencers can make political messaging feel more authentic and conversational
- some political influencer work happens without clear disclosure that money changed hands
Main risks and flaws
Problems with chatbots
The guests highlight several issues with using AI as a voting guide:
- Hallucinations: bots can make things up or confidently repeat falsehoods
- Bias from sources: they may pull from non-neutral or incomplete information
- Persuasion by design: bots are built to be agreeable and keep users engaged
- False confidence: users may feel informed even if the chatbot simply reinforced what they already believed
The episode emphasizes that a chatbot can appear neutral because it lacks human personality, but that absence of visible bias does not make it objective.
Problems with influencer politics
Riedel argues that the influencer ecosystem has major transparency issues:
- there are few norms or ethical guardrails
- disclosure rules are often weak or nonexistent
- paid political content can look like genuine grassroots support
- dark money can flow through creator campaigns with little accountability
He compares this dynamic to astroturfing: manufactured support that appears organic.
How to use these tools more carefully
Advice for chatbot users
Javed recommends:
- asking neutral, non-leading questions
- avoiding personal accounts or using incognito mode
- staying skeptical about where the chatbot’s information comes from
- asking who owns the platform and what its motives might be
Advice for media consumers
Riedel suggests being aware that:
- many people assume creators are being paid anyway
- the real concern may be informing less-online users, who are less likely to recognize sponsored political content
- disclosure matters even if heavy social media users have normalized paid creator content
Bigger takeaway
The episode argues that the shift toward a “chronically online election” reflects a broader change in how people get political information. Voters are moving away from legacy media and toward chatbots, social feeds, and creator-driven content. Whether that is a healthy evolution or a dangerous one depends on transparency, media literacy, and whether campaigns and platforms take responsibility for the accuracy and disclosure of what they put in front of voters.