Using AI To Help Research Online Anonymity and Political Speech
Introduction
During the 2016 US elections when Donald Trump commented that Russian hacking of Democratic National Committee emails was less likely than “…somebody sitting on their bed that weighs 400 pounds,” the US intelligence community was already aware of foreign attempts to influence US elections.
My own take on the situation was that the prevailing allowance by US publications of anonymous comments was one of the most addressable issues, i.e., if US publications were to do away with anonymous commenting, this would lead to less disinformation and better (and more civil) online discussions.
However, as I noted in 2023’s Some Folks Just Don't Want to Pull Back the Identity Curtain, not everyone supports online anonymity. The usual reason given in response to my own belief is that anonymity protects online commenters from retribution for unpopular opinions. This rationale is supplied both by those on the “left” as well as those on the “right” and suggests to me at least that there might be some loss to free opinion expression were online anonymity discontinued.
On the other hand, there might be a reduction in disinformation and outright deception if individuals were required to announce their identities online. So, how can the arguments pro and con regarding online commenting anonymity be balanced? My first step was to summarize the arguments pro and con and to do that I turned to 3 online AI based search tools.
Posing the question
On May 28, 2024 I asked three different online AI tools (Claude, Perplexity, and ChatGPT-4o) to summarize the arguments for and against allowing anonymous comments about political topics in online publications. I then asked ChatGPT-4o to synthesize the results of the three searches. Below is ChatGPT-4o’s synthesis of the arguments for and against allowing anonymous comments:
Summary of Arguments Supporting Allowing Anonymous Comments on Political Topics:
Protection and Safety: Anonymity allows people to express controversial or dissenting views without fear of retaliation, harassment, or social stigma, and is crucial for the safety of commenters in oppressive regimes.
Freedom of Speech: It protects individuals' rights to freely express their opinions, especially in politically sensitive contexts, supporting a more open discourse.
Increased Participation: Anonymity encourages greater participation and engagement from individuals who might otherwise remain silent due to fear of judgment or reprisal.
Equal Playing Field: It levels the playing field by focusing discussions on the content of arguments rather than the identities or backgrounds of the commenters.
Diverse Perspectives and Whistleblowing: Anonymity enables a wider range of voices and perspectives, including those of whistleblowers and insiders, enriching the debate and leading to a more comprehensive understanding of political issues.
Summary of Arguments Opposing Allowing Anonymous Comments on Political Topics:
Reduced Accountability and Civility: Anonymity often leads to a lack of accountability, enabling abusive language, hate speech, threats, trolling, and other uncivil behaviors that commenters would be less likely to engage in if identifiable.
Manipulation and Misinformation: Anonymous comments are more susceptible to manipulation through coordinated campaigns, bots, and foreign interference, making it difficult to trust the narratives being presented.
Credibility Issues: The lack of attribution makes it challenging to assess the credibility, expertise, and potential biases of commenters on complex political issues, undermining the quality of discourse.
Harassment and Toxic Behavior: Anonymity enables harassment, trolling, flaming, and "dogpiling" on targeted individuals without consequences, leading to a toxic online environment.
Degradation of Discourse Quality: Anonymous comments tend to be more uncivil, containing insults, stereotyping, and factual inaccuracies, which degrade the overall quality of discourse and reduce trust in journalism and media outlets.
Researching the question
As a longtime management consultant, it’s natural for me to look at different options in terms of costs and benefits, each having both quantifiable and non-quantifiable aspects.
The most quantifiable aspect of online publications doing a better job of validating the identity of people desiring to make public comments would be the cost of background research: are these people who they say they are, and what kind of information should we publish and maintain online to allow readers some idea of the qualifications of the person making the comments?
Usually when I complain about online anonymity I get two pushbacks from commenters:
Fear of retribution from those with opposing views.
Loss of revenue by the online publisher due to a reduction in measurable online engagement.
Both of these are represented in the above AI-based analysis and I think have to be taken into account given today’s divided political environment and the ever-precarious finances and advertising-dependence of online publications.
Were I to continue this avenue of research my next step would be to take each of the above “pro” and “con” lists and dig deeper into each one. As an experiment I asked ChatGPT-4o the following question:
Please outline a research project to test the following hypothetical arguments concerning support or opposition to allowing anonymous comments about political topics in online publications:
I then pasted in the two lists of pro and con arguments, and this is how ChatGPT-4o responded:
Research Project Outline: Testing Support or Opposition to Allowing Anonymous Comments on Political Topics in Online Publications
1. Research Objectives
To evaluate the validity of hypothetical arguments supporting and opposing the allowance of anonymous comments on political topics.
To assess the impact of anonymous commenting on the quality, civility, and diversity of political discourse.
To analyze the relationship between anonymity and participation rates in online political discussions.
2. Research Questions
How does anonymity affect the expression of controversial or dissenting views in political discussions?
Does anonymity encourage greater participation and engagement in online political discourse?
What is the impact of anonymity on the civility and accountability of comments?
How does anonymity influence the diversity of perspectives and the prevalence of whistleblowing in political debates?
How do anonymous comments affect the quality and credibility of political discourse?
3. Hypotheses
Supporting Allowing Anonymous Comments:
Anonymity increases the expression of controversial or dissenting views.
Anonymity boosts participation and engagement in political discussions.
Anonymity leads to a more diverse range of perspectives, including whistleblowing.
Opposing Allowing Anonymous Comments:
Anonymity reduces accountability, leading to more uncivil behavior and hate speech.
Anonymous comments are more susceptible to manipulation and misinformation.
Anonymity undermines the credibility and quality of political discourse.
4. Methodology
Study Design:
Mixed-methods approach combining quantitative surveys, content analysis, and qualitative interviews.
Participants:
A diverse sample of online commenters, readers, and experts in political communication, journalism, and digital ethics.
Data Collection:
Surveys:
Conduct surveys with online commenters to measure their views on anonymity, safety, freedom of speech, and participation.
Use Likert scales to assess perceptions of civility, accountability, and discourse quality.
Content Analysis:
Analyze comments from multiple online political forums and news sites, comparing anonymous and attributed comments.
Measure variables such as civility, hate speech, diversity of perspectives, and instances of whistleblowing.
Qualitative Interviews:
Conduct in-depth interviews with frequent commenters, moderators, and experts to gain insights into the motivations, experiences, and perceived impacts of anonymous commenting.
Experimental Design:
Create controlled online discussion forums with varied commenting conditions (anonymous vs. attributed).
Measure participation rates, quality of discourse, civility, and diversity of perspectives across different conditions.
5. Data Analysis
Quantitative Analysis:
Use statistical techniques to compare survey responses and comment analysis results between anonymous and attributed comments.
Employ regression analysis to determine the impact of anonymity on participation, civility, and discourse quality.
Qualitative Analysis:
Thematic analysis of interview transcripts to identify common themes and insights related to the impacts of anonymity on political discourse.
6. Ethical Considerations
Ensure participant anonymity and confidentiality in surveys and interviews.
Obtain informed consent from all participants.
Address potential biases in data collection and analysis.
7. Expected Outcomes
Insights into the benefits and drawbacks of anonymous commenting on political topics.
Recommendations for online platforms and policymakers on managing anonymous comments to balance freedom of expression with the need for civility and accountability.
Enhanced understanding of the role of anonymity in promoting diverse perspectives and whistleblowing.
8. Dissemination of Findings
Publish findings in academic journals and present at conferences related to political communication, journalism, and digital ethics.
Share results with online platforms and media organizations to inform policy and moderation practices.
Conduct public webinars and workshops to engage with a broader audience on the implications of anonymous commenting.
This research project aims to provide a comprehensive evaluation of the arguments for and against allowing anonymous comments on political topics, contributing to a more informed debate on this contentious issue.
I’m underwhelmed
To be honest, I was severely underwhelmed by this research plan. While it might serve as a good jumping off place to do “meatier” project planning, I couldn’t help but think I might have come up with a standard outline like this back when I was in grad school.
Based on other research on AI and project planning, however, I believe that a more detailed statement of research methodology requirements in the initial prompt would generate a more focused—and realistic—plan outline. Maybe I’ll try that next.
What else is missing?
The major thing missing from this plan outline is direct attention to what I believe increasingly is a serious problem with online communication: bad actors who explicitly use disinformation and lies to deceive. For example, how can I know if a commenter on a political article in The Washington Post is a Democrat, a Republican--or a Russian supporting Russian interests?
The above research outline does not directly address questions of foreign influence and intentional deception. If publications allow anonymous comments I will continue to view such comments as more entertaining than informational.
Copyright © 2024 by Dennis D. McDonald. The image at the top of the page was generated by Microsoft Designer Image Creator in response to the prompt, “Please create a line drawing illustrating how the issue of online anonymity can be researche using social science methods.”