AI can help voters research candidates, but it should support—not replace—independent civic judgment.
CIVPAC would like to have the resources to provide endorsements in the vast majority of primaries and general-election contests. We do our best, but sadly that goal is, at least for now, beyond our reach.
If you find our philosophy and policy positions similar to your own, and you do not find an endorsement from us in a race you care about, we do have a suggestion. You can use your favorite AI tool to assist you in getting to a decision.
We suggest copying and pasting the following prompt into your AI tool. Within the prompt, replace the words [INSERT RACE] with the race you are interested in. Of course, you may want to modify it to reflect your own preferences.
Suggested AI Prompt
“I am a centrist voter broadly aligned with CIVPAC-style principles: market-compatible solutions; fiscal seriousness; equality of opportunity; respect for personal freedom; practical liberal-democratic governance, including respect for constitutional limits, democracy, and the rule of law; support for Ukraine and stable international alliances; and skepticism toward both far-right authoritarian populism and far-left or DSA-style politics.
Please help me evaluate the candidates in this race: [INSERT RACE].
Do not simply tell me who to vote for. Instead:
- Identify the major candidates and the election context.
- Summarize each candidate’s record, platform, endorsements, major public statements, and relevant controversies.
- Compare the candidates against the principles above.
- Flag evidence of election denial, authoritarianism, corruption, contempt for constitutional limits, antisemitism, isolationism, hostility to NATO or Ukraine, unserious fiscal promises, or extreme ideological commitments.
- Distinguish clearly between facts, reasonable inferences, and opinion.
- Use current, reliable sources and provide citations or links.
- Explain which candidate appears most consistent with these principles and why.
- Identify serious weaknesses, uncertainties, or gaps in the evidence.
- If this is a primary, analyze the choice within that party’s field and do not assume that my primary preference will be the same as my general-election preference.
- Keep the analysis practical. Perfect ideological alignment is too much to ask. Governing temperament, institutional responsibility, policy seriousness, and ability to represent the district can make up for some degree of policy weakness in a candidate.”
After You Get the AI Response
Do not treat the AI’s answer as a substitute for your own judgment. AI tools can be useful research assistants, but they can miss recent developments, rely on stale sources, overstate uncertain conclusions, or fail to understand local political context. Ask for sources. Check the most important claims. Then make your own decision. AI engines can have subtle hidden biases. Press your tool to be more specific if its reasoning is not compelling for you.