The rise of artificial intelligence (AI) has sparked a revolution in various industries, including politics. While AI can be perceived as an unstoppable force shaping society, it is crucial to recognize that humans can have control over how we use […]
The rise of artificial intelligence (AI) has sparked a revolution in various industries, including politics. While AI can be perceived as an unstoppable force shaping society, it is crucial to recognize that humans can have control over how we use these tools.
As political scientists, we see these new technologies as an opportunity to enhance democratic processes, provided that we address certain challenges. We have begun researching how this can be achieved.
Increasing Citizen Awareness
Politics can appear overwhelmingly complex, with negative marketing campaigns and a constantly shifting political landscape. Many cities, states, and countries provide little to no information to the public about political issues, candidates, or referendums. Consequently, even when citizens have the opportunity to exercise their democratic freedoms, they may not feel sufficiently informed to do so.
Generative AI can help. Through platforms like isidewith.com, politicalcompass.org, and theadvocates.org, AI can assist people in answering questions about their core beliefs or political stances and help them determine which candidates, parties, or elections align best with their views.
Existing websites like Ballotpedia, Vote Smart, and Vote411 have made significant strides in providing key information such as sample ballots, polling locations, or candidate positions to voters. However, these websites can be difficult to navigate. AI can provide enhanced services at the local, state, regional, national, and international levels. These systems can potentially use automation to provide continuously updated information on candidates and political issues.
AI chatbots can also interactively help individuals understand complex issues, learn new skills, and define their political opinions by providing relevant news and facts.
However, current generative AI systems are not yet ready to reliably answer democracy-related questions without bias or accuracy. Large language models generate text based on statistical word frequencies from their training, regardless of whether statements are true or fictional.
For example, AI systems can “hallucinate” by creating fictional politicians or generating inaccurate candidate positions. These systems also exhibit political biases in their output. Rules that protect user privacy and provide compensation to individuals or organizations whose data is used by these systems are still unclear.
Before generative AI can truly enhance democracy, there are many things that need to be understood and resolved.
Facilitating Communication with Representatives
One area that requires investigation is whether generative AI can help citizens communicate with their elected representatives.
Contacting politicians can be intimidating, and many individuals may not even know where to begin. Surveys have revealed that less than half of Americans can name the three branches of government. Knowing the names of their own representatives, let alone reaching out to them, is even less common. For instance, only 23% of respondents in a 2018 Pew Research Center survey stated they had contacted an elected official in the past year, even during significant national events.
To encourage greater citizen engagement in communication with representatives, generative AI can assist individuals not only in identifying their elected officials but even in drafting detailed letters or emails to them.
We explored this idea in a recent study conducted within our work at the Laboratory for Responsible AI and Management at Purdue University. We conducted a survey among adult Americans in June 2023 and found that 99% of respondents were aware of generative AI systems like ChatGPT, and 68% had personally tried them. However, 50% stated they had never contacted any political representative.
Within the survey, we showed some respondents an example message written by ChatGPT to a state legislator regarding an education funding bill. The control group saw the same email example, but without indication that it was generated by AI.
Respondents who were aware of this potential use of AI technology expressed significantly greater support for using AI to communicate with politicians, both by individuals and advocacy groups. Due to their support for this new technology, we expected them to express a greater inclination to contact politicians more frequently and consider AI as facilitating that process. However, they were not inclined to do so.
Nevertheless, we have identified an opportunity. For example, public interest groups could leverage generative AI to enhance advocacy campaigns by helping citizens personalize their emails to politicians more easily. If they can ensure that the messages generated by AI are fact-based and accurately reflect citizens’ views, many more people who might not otherwise contact their politicians could consider doing so.
However, there are risks, including politicians’ suspicion that messages are written by AI.
1. What is generative artificial intelligence (AI)?
Generative AI is a type of artificial intelligence that has the ability to produce original and creative outputs, such as text, images, or sounds.
2. How can generative AI assist political engagement?
Generative AI can assist political engagement by enabling citizens to easily access information about political topics, candidates, and elections. It can also facilitate communication between citizens and elected representatives through personalized messages and letters.
3. What are the advantages and challenges of using generative AI in political engagement?
The advantages of using generative AI in political engagement include increased citizen awareness, personalized communication with politicians, and easier access to critical information. However, challenges include the reliability of AI-generated outputs, political biases, and issues of data privacy.
4. What are the risks in using generative AI in political communication?
Risks include politicians’ suspicion of AI-generated messages, potential political biases in generated outputs, and unclear rules regarding data privacy protection.