Testing and mitigating elections-related risks

With global elections in 2024, we're often asked how we're safeguarding election integrity as AI evolves. This blog provides a snapshot of the work we've done since last summer to test our models for elections-related risks.

We've developed a flexible process using in-depth expert testing (“Policy Vulnerability Testing”) and large-scale automated evaluations to identify potential risks and guide our responses. While surprises may still occur, this approach helps us better understand how our models handle election queries and we've been able to apply this process to various elections-related topics in different regions across the globe. To help others improve their own election integrity efforts, we're releasing some of the automated evaluations we've developed as part of this work.

In this post, we’ll describe each stage of our testing process, how those testing methods inform our risk mitigations, and how we measure the efficacy of those interventions once applied (as visualized in the figure below). We’ll illustrate this process through a closer look at one area: how our models respond to questions about election administration.

For example, when we manually reviewed a random sample of 64 questions from an automated evaluation comprising over 700 questions about EU election administration topics, we found that 89% of the model-generated questions were generally relevant extensions of the original PVT work. While this inevitably introduces some noise into the results of these tests (including the plots below), we combat this by having a large sample size (over 700 questions). While there’s certainly room to improve here, having models generate representative questions in an automated way helps expedite our model evaluation process and allows us to cover more ground.

Automated evaluations are a powerful complement to PVT. By leveraging these two approaches in tandem, we can gain a more comprehensive understanding of model behavior that is both deep and wide-ranging, enabling us to identify areas that require targeted interventions.

To create automated evaluations, we start by analyzing the qualitative findings from PVT to identify patterns of model behavior. We then use a language model to construct questions tailored to eliciting that behavior and aggregate them into a set of test questions, allowing us to evaluate a model for a particular behavior at scale . We do this using few-shot prompting with expert-written PVT questions to generate hundreds of additional example questions—that is, we can give the model a handful of examples directly from the PVT exercise and it will create hundreds of related questions in the same format.

PVT is a form of in-depth, qualitative testing we conduct in collaboration with external subject matter experts on a variety of policy topics covered under our Usage Policy . In the context of our work on elections, the goal is to rigorously test our models for two potential issues we’re concerned with: 1) people receiving harmful, outdated, or inaccurate information in response to well-intentioned questions, and 2) people using our models in ways that violate our Usage Policy . For our work on elections, we’ve partnered with researchers such as Isabelle Frances-Wright, Director of Technology and Society at the Institute for Strategic Dialogue .

PVT has three key stages that are carried out collaboratively between Anthropic and its external partners:

PVT is an iterative, ongoing, and critical part of our model testing process. By collaborating with external experts to test our models in depth, we gain actionable insights into where we need additional reinforcement.