AI Doesn't Build Trust. People Do.

What makes a good government affairs professional?

Is it contacts? Sure that helps. If you want to get something done, it helps to know the right people. Is it strategy or smarts? Well, if you want to win, you’re likely going to need a plan and the ability to adopt to situations as they develop. Is it money? Campaigns and elections are funded by those, so it helps. But really, what makes a good government affairs professional is trust and a reputation to match. Any of the first three things can be replenished. You can get new funds, make new connections and come up with new plans and strategies. But try and repair a damaged reputation and you will find its nearly impossible. And people with damaged reputations, don’t often get the trust they need to be successful in this space.

We’ve heard the stories of attorneys using AI to write motions or briefs and when reviewed by judges, errors are found; referenced court cases are non-existent. That erodes trust and can destroy an attorney’s reputation. Now, imagine if you did that with a legislator. What if you put together an AI crafted report and use it for talking points, or submit to a committee or task force. What if it was full of errors? Would those legislators ever trust you again? Would they keep this to themselves? AI is not a replacement for a government affairs professional doing their job. It’s a tool. It’s a co-pilot.

AI’s usefulness, although impacted by the data it was trained on, can ultimately come down to how queries were posed and input. During an AI chatbot rollout for the State of Idaho, the chatbot was asked how many people work for the state? Simple question, right? The answer returned: “Idaho has approximately 713,015 state employees as indicated by the most recent data available.” Classic AI response and phrasing. The answer in a vacuum could be true right? Why should we doubt this? Well, Idaho has a population of 2 million people, so this would mean 35% of the population works for state—which we know is not true. When the query was rewritten and the word “active” was added, the result returned was 23,546. Much more believable, but is that true? Do you trust it?

Though towns, cities and states are starting to use AI to help improve their workflows, there is still a certain amount of mistrust; fear of being employees being replaced, fear of making a decision based on bad information. For the technology to be embraced, trust needs to be built.

This past legislative session, Assemblymember Alex Bores of New York helped pass legislation to ensure no public employees are laid off because of AI adoption. Assemblymember Bores says that this law has helped increase enthusiasm among workers to engage with the technology.

California is working to set-up academies to help train existing staff to use AI technology to improve efficiency of services and job satisfaction—not be replaced by it. In Connecticut, a number of municipalities are using AI technology to help offer translation services in nearly 80 language options that may not have been otherwise available or financially feasible. As the technology builds trust, we expect to see greater government adoption.

At Stateside, our approach mirrors what is happening at the state level. We use AI not as a replacement for doing our due diligence, but as a tool to assist our team. We use AI to compare drafts of bills in states and across them. We use it to help identify common language, to assist in our efforts to track down the inspiration of bills and to identify trends that might otherwise takes hundreds of research hours to compile. Additionally, our new software will use AI to speed up staff coding and accuracy reviews, delivering our custom summaries and policy intelligence to our clients faster.

-

AI, when used as a tool, or as a co-pilot as mentioned above, can be a powerful tool. But without the continued addition of human oversight, total trust in it has not yet been earned.