When Donors Ask If Your Emails Are Written by AI

Sooner or later a donor is going to ask whether the candidate actually wrote the email they just got. It can come from someone who has given for years and noticed the emails got sharper, from a volunteer who knows how campaigns run, or from a reporter working a story about AI in this election.

That last one is not far-fetched this year. On September 25 the Associated Press reported that the Wesleyan Media Project has counted at least 164 political ads made or enhanced with AI this cycle, and that 69% of them never said so. Donors read those stories, and some of them are going to look at their inbox and wonder.

So have the answer ready before anyone asks.

The honest answer

"AI helped draft it. A person on our team reviewed it, the positions in it are the candidate's, and nothing went out until someone approved it."

That is the whole answer, and it only works if every word of it is true. If your process can't support that sentence, fix the process before you worry about the wording.

How you say it matters as much as what you say. Lead with it apologetically and it sounds like a confession. Lead with how the work actually gets done and it sounds like a campaign that has its act together, because every donor is getting a real, reviewed note built from their own history with the campaign.

This is how campaign writing has always worked

Candidates have never written most of what goes out under their name. The press release comes from the communications director, the fundraising email comes from a staffer working off last cycle's best performer, and the talking points come from whoever knows the issue cold. Nobody who has spent time around a campaign expects anything else.

What changed is who writes the first draft. What didn't change is who is accountable for it: the campaign, and the person who approved it before it went out.

The question underneath the question

Underneath "did AI write this?" there is a better question, and it's the one worth being ready for: does this campaign actually know who I am, or am I just on a list?

The answer depends on how the email was built. If the draft came from the donor's real history with the campaign, the notes from the last conversation, and the candidate's actual position on the issue, then yes, the campaign knows them, and the email is the proof. If it came from a list and a template, the donor has caught you, and it doesn't matter whether a person or a machine filled in the blanks. A better answer about software won't fix that.

When it comes up in person

The version that takes the most preparation is the one at an event, when a donor pulls the candidate or the finance director aside and asks straight out. Have these ready.

Don't get defensive. "Yes, we use AI to help draft our emails" is a complete sentence, and it is not a scandal.

Go straight to the approval. The thing to settle is that a person read it before it went out. Say that first and the rest of the conversation gets easy.

Tie it to capacity. AI lets a small team answer every donor personally, instead of deciding which donors are big enough to get a real note.

Keep it short. One confident paragraph does more than a lecture on AI policy.

What the law says right now

In California, the disclosure rule is AB 2355, in effect since January 1, 2025. It requires a political ad that uses AI-made or AI-altered images, audio or video, the kind a reasonable person could mistake for real, to carry the label "Ad generated or substantially altered using artificial intelligence." It is aimed at fake media: the doctored photo, the cloned voice.

At the federal level, the Federal Election Commission decided in September 2024 not to write new AI rules. Instead it said its existing ban on fraudulently misrepresenting who you are applies no matter what technology you use.

Other states have passed their own AI ad laws, and they don't all say the same thing. If an email carries an AI-made image or video, or your committee is active in more than one state, run it past your campaign attorney before it goes out.

Whether to say so in the footer

Some campaigns will add a line to their email footer saying AI helps with drafting and a person approves every message. The case for it is that it's accurate and it answers the question before anyone asks. The case against it is that it can raise a question some donors were never going to ask.

If your process is clean, either choice is defensible. What isn't defensible is a process you would be embarrassed to describe out loud.

The part that matters

The AI isn't the risk. The risk is a campaign that lets a machine say something the candidate never said, or sends something nobody read. Build it so a person approves everything, and when a donor asks, the true answer is also the best one you could give.

That is how we build it at AutomatedTeams: the machine drafts, a person releases it, and nothing goes out in the candidate's name without a yes. Here's how an engagement works.

Eric Linder served in the California State Assembly (60th District, 2012 to 2016) and founded AutomatedTeams, a political AI consulting firm for campaigns, committees, PACs and advocacy groups.