Second, we need to strengthen disclosure requirements for lobbyists, whether they are fully human or AI-powered. State laws regarding lobbying disclosures are a quagmire. North Dakota, for example, only requires annual filing of lobbying reports, so by the time the disclosure is made, the policy is likely to have been decided. A lobbying disclosure scorecard produced by Open Secrets, a group that studies the influence of money in US politics, tracks nine states that don’t even require lobbyists to report their compensation.
Ideally, it would be great for the public to see all communications between lobbyists and lawmakers, whether it takes the form of a proposed amendment or not. Failing that, let’s give the public a chance to consider what lobbyists lobby – and why. Lobbying is traditionally an activity that takes place behind closed doors. Many states now reinforce this, effectively exempting testimony given publicly to the legislature from being considered lobbying.
In these jurisdictions, if you disclose your position to the public, you are no longer lobbying. Let’s do the opposite: require lobbyists to reveal their positions on issues. Some jurisdictions already require a position statement (“yes” or “no”) from registered lobbyists. And in most (but not all) states, you can make a public records request regarding meetings held with a state legislator and hope to get something substantial back. But we can expect more — lobbyists can be required to proactively publish, within days, a summary of what they have demanded from politicians during meetings and why they believe it is in the public interest.
We cannot expect corporations to be open and completely honest about the reasons for their lobbying positions. But the fact that they communicate their intentions will at least be a basis for accountability.
Finally, consider the role of AI-assisted technologies for lobbying companies themselves and the labor market for lobbyists. Many observers are rightly concerned about AI’s potential to replace or devalue the human work it automates. If the automation potential of artificial intelligence eventually commoditizes the work of developing policy strategies and crafting messages, it could indeed put some professionals on Kay Street out of a job.
But don’t expect it to derail the careers of the most astronomically compensated lobbyists: former members of Congress and other insiders who come through the gate. There is no shortage of ideas for reforms to limit the ability of government officials turned lobbyists to sell access to their colleagues still in government, and they should be adopted and — just as importantly — supported and enforced by successive Congresses and administrations.
None of these solutions are truly original, specific to the threats posed by AI, or even predominantly focused on micro-regulation – and that’s the point. Good governance must and can be resilient to threats from different methods and actors.
But what makes the risks associated with artificial intelligence particularly relevant now is how quickly the field is evolving. We expect the scale, strategies and effectiveness of lobbying people to evolve over years and decades. Meanwhile, progress in artificial intelligence seems to be making impressive breakthroughs at a much faster pace—and it’s still accelerating.
The legislative process is a constant battle between parties trying to control the rules of our society as they are updated, rewritten, and expanded at the federal, state, and local levels. Lobbying is an important tool for balancing different interests through our system. If it is well regulated, perhaps lobbying can support politicians to make fair decisions on behalf of us all.
Nathan E. Sanders is a data scientist with the Berkman Klein Center at Harvard University. Bruce Schneier is a security technologist, fellow, and faculty member at the Harvard Kennedy School.
