Participating in elections, whether by voting or running for office, is the lifeblood of democracy.
The harder it is (or the harder it appears) for individuals to make their voices heard, the less agency they feel they have and the more our institutions get chilled. If it’s too costly or inconvenient to get involved, why bother?
In 2020, the average cost of running a successful campaign for the US House was $2 million. A successful US Senate campaign cost on average $20 million over the same period, according to an analysis by the Campaign Finance Institute. Running for a state senate seat on average cost around $120,000, and even running for a school board seat can exceed $25,000.
That’s a huge barrier for someone contemplating a first campaign. Anything that trims these costs by even a little can make running more approachable.
This is where artificial intelligence comes in.
Much attention has been paid to the threats of disinformation and misinformation emanating from the technology, and rightly so. Last month an attempted deepfake operation against Senate Foreign Relations Chair
But this shouldn’t distract from the technology’s very pragmatic potential to reduce the kinds of mundane but necessary costs that help make campaigns so expensive. Further, when deployed in the right way, AI can diversify financial support for political candidates and intensify direct appeals to supporters. Grassroots projects have used these to great effect through the use of social media.
As my colleagues Oma Seddiq and Zach Cohen reported in August, some campaigns have begun using various off-the-shelf products to perform pedestrian tasks. AI may not be able to craft an emotive stump speech or determine real-time campaign strategy, but it can free up time for door knocking or honing a complex policy message.
Progressive campaigns, for example, can sign up to use a tool from Quiller AI for $20.24 per month that can help draft emails, digital ad copy, op-eds and press releases.
Truverse, meanwhile, uses chatbots to streamline how political operatives interact with voters, much in the same way online retailers use the technology to streamline customer service.
While the savings might be a drop in the ocean for congressional campaigns with multi-million-dollar budgets, they can make a real difference for state and local candidates fighting for every dollar.
Netflix subscription-like tools that streamline the technics of running for office also provide a cheap way to test the effectiveness of messaging. Someone eyeing a school board seat can—at the click of a button—test if their message resonates (and whether a strategy or campaign is worth pursuing).
We expect an instant response from the technology we use in our everyday lives, and these tools can offer the same to political candidates.
Of course, whether the technology becomes a net positive for the political process will depend on whether lawmakers and the executive branch can create clear guardrails that rein in the dissemination of falsehoods while protecting freedom of expression. It remains to be seen if recent discussions on a potential year-end AI bill bear fruit during the lame-duck session of Congress in December.
Who has access to these tools is also a central question. Many of the tools backed by Silicon Valley investment are available only for use by progressive campaigns. As technologists grapple with wider questions about AI bias, what does it mean when the tools used by political operatives are trained on datasets that lead to certain outcomes?
But if AI helps get us a broader slate of candidates with clearer messages and expanded coalitions of donors and supporters, that’s not bad for democracy.
To contact the reporter on this story:
To contact the editors responsible for this story: