Politicians Are Trying to Change What Chatbots Say About Them
As AI chatbots become a front door to information, political candidates and officials are pressuring developers to shape how models describe them.
In Brief
- Politicians and campaigns are working to shape what AI chatbots say about them
- Tools now let campaigns see and influence what LLMs tell voters
- The pressure ranges from persuasion attempts to legal threats against chatbot makers
As AI chatbots become a front door to information for millions of voters, politicians and campaigns have started working to shape exactly what those models say about them — turning the answers of ChatGPT and Claude into a new battleground.
The dynamic was documented by The New York Times in a piece on politicians trying to change what chatbots say about them. The effort is no longer hypothetical: campaigns are treating LLM answers as a channel to be managed like any other.
Other outlets confirm the trend. NPR reports political campaigns are turning to AI tools to text and engage voters, with one organizer noting the risk of generative AI “outweighs the benefits, including the chatbot giving false information.”
Notus reports that some groups “have said publicly that they’re trying to shape what answers LLMs give to people who ask about politics,” an effort it compares to a supercharged version of long-standing campaigns to influence coverage.
How campaigns reach the chatbots
New tooling is making YouTube’s AI content monetization crackdown practice concrete. Reports describe software that lets campaigns see what chatbots are telling voters about a candidate, then adjust websites, statements, and talking points to steer those answers — the same SEO logic applied to generative models.
The methods range from legitimate to coercive. Some campaigns simply publish clearer, model-friendly information; others, per Platformer, involve legal pressure, as when Missouri’s attorney general launched a campaign arguing chatbot makers are censoring certain political figures.
The result is a gray zone. Unlike a search engine result or a news article, a chatbot answer feels like a confident judgment, which gives any successful manipulation outsized weight with undecided users. OpenAI Has Plan
Why this matters for information integrity
Chatbots are rapidly becoming a default source for quick factual lookups, which means the “front page” of AI answers can shape elections as much as traditional media once did. Whoever influences those answers gains a quiet advantage.
The risk is asymmetry and opacity. Voters rarely know a campaign has optimized for an LLM, and the model itself may not disclose the provenance of its stance. That undermines the promise of chatbots as neutral, factual assistants.
For the model makers, the challenge is governance: how to resist coordinated attempts to rewrite political answers without themselves becoming arbiters of truth. The politicians-and-chatbots story is only beginning, and the rules are not yet written.
FAQ
Are politicians really trying to influence chatbots?
Yes. The New York Times and other outlets report campaigns and officials are working to shape what ChatGPT, Claude, and other chatbots say about them.
How do campaigns influence chatbot answers?
They use tools that reveal what models tell voters, then adjust public information and talking points; some also apply legal pressure on model makers.
Why does this matter for elections?
Chatbots are becoming a primary source of information for voters, so influencing their answers can quietly shape political outcomes.
