ECIS 2026 (2026) AI Processed Human Approved
GENAI AS A ROMANTIC RELATIONSHIP ADVISOR:EXPLORING HUMAN–AI COLLABORATION IN EMOTIONAL SUPPORT CONTEXTS
This qualitative study investigates how individuals use Generative AI (GenAI) as a reflective and preparatory tool for romantic relationship challenges. Using the Human-AI Interaction framework (HAII-TIME) as an analytical lens, the researchers conducted 18 semi-structured interviews in 2025 with GenAI experts, professional psychologists, and end users to capture different perspectives on AI-assisted relationship guidance.
Problem
Although GenAI tools are increasingly accessed as low-threshold, private, and immediate sources of relationship advice, little is known about the relational consequences and risks in emotionally sensitive contexts. Traditional counseling relies on multi-perspective, balanced dialogue, whereas individual GenAI interactions are highly vulnerable to one-sided framing and lack key nonverbal contextual cues.
Outcome
- Qualitative evidence supported that GenAI helps users self-clarify emotions, organize thoughts, and prepare communication strategies prior to interpersonal discussions.
- Qualitative evidence supported that GenAI use entails risks of automation and confirmation bias, as the system tends to validate the user's one-sided, potentially biased framing of relationship conflicts.
- Qualitative evidence supported that users often attribute unintended authority to GenAI, potentially misinterpreting its structured, polished responses as clinical diagnoses.
- Qualitative analysis supported that GenAI is a useful complementary or preparatory tool but cannot replace professional therapy because it lacks emotional intuition, nonverbal communication perception, and dyadic mediation capabilities.
- No statistical hypotheses were tested due to the study's exploratory qualitative nature; findings are limited by a small sample size of 18 interviewees, retrospective self-reports, a German-speaking context, and homogeneous use of ChatGPT.
- Qualitative evidence supported that GenAI use entails risks of automation and confirmation bias, as the system tends to validate the user's one-sided, potentially biased framing of relationship conflicts.
- Qualitative evidence supported that users often attribute unintended authority to GenAI, potentially misinterpreting its structured, polished responses as clinical diagnoses.
- Qualitative analysis supported that GenAI is a useful complementary or preparatory tool but cannot replace professional therapy because it lacks emotional intuition, nonverbal communication perception, and dyadic mediation capabilities.
- No statistical hypotheses were tested due to the study's exploratory qualitative nature; findings are limited by a small sample size of 18 interviewees, retrospective self-reports, a German-speaking context, and homogeneous use of ChatGPT.
What it means for you
- CIO / IT Executive: Configure the enterprise GenAI gateway to detect personal counseling or emotional-support-seeking queries and automatically inject a system-level disclaimer redirecting employees to the company's official Employee Assistance Program (EAP) to prevent reliance on GenAI as an authoritative clinical resource.
- IT Manager: Create and distribute a 'Balanced Prompting' template to the team, instructing them to explicitly prompt GenAI to 'act as a devil's advocate and challenge my assumptions' when using it to prepare for difficult workplace or interpersonal conversations, directly mitigating the tool's confirmation bias.
- Business Strategist: Review the product roadmap for our wellness and conversational AI features to pivot the UX away from direct 'advice-giving' toward 'communication preparation and self-reflection' tools, integrating mandatory disclaimers that clarify the AI has no clinical diagnostic capabilities.
- Researcher: Draft a research proposal for a pre-registered quantitative study ($N > 300$) comparing how different LLM models (e.g., ChatGPT, Claude, Gemini) handle relationship conflict prompts, specifically measuring the rate of user agreement (automation bias) versus the model's level of validation.
- Policymaker: Draft a regulatory policy mandate requiring developers of consumer-facing GenAI tools to implement prominent 'non-clinical tool' labels and active-consent pop-ups warning users of the risks of automation bias and the lack of professional diagnostic authority in emotionally sensitive domains.
Transcript
Host: Welcome to A.I.S. Insights - Turning IS Research into Business Action. I am your host, Anna Ivy Summers, and today we are discussing a study titled "GENAI AS A ROMANTIC RELATIONSHIP ADVISOR: EXPLORING HUMAN–AI COLLABORATION IN EMOTIONAL SUPPORT CONTEXTS." This qualitative study investigates how individuals use Generative AI as a reflective and preparatory tool for romantic relationship challenges. Joining me to break down this research is our expert analyst, Alex Ian Sutherland. Welcome, Alex.
Expert: Thank you, Anna. It is great to be here to discuss this study. The researchers used the Human-AI Interaction framework, known as HAII-TIME, as an analytical lens. They conducted 18 semi-structured interviews in 2025 with GenAI experts, professional psychologists, and end users to capture different perspectives on AI-assisted relationship guidance.
Host: Let's start with the real-world problem this study addresses. Why are people turning to Generative AI for relationship advice in the first place?
Expert: Well, traditional counseling can be difficult to access due to waiting times, costs, scheduling, or social stigma. GenAI tools offer a low-threshold, private, and immediate alternative. However, the problem is that we know very little about the relational consequences and risks when these tools are used in emotionally sensitive contexts. Traditional therapy relies on a balanced, multi-perspective dialogue, whereas individual interactions with GenAI are highly vulnerable to one-sided framing and lack nonverbal cues like body language.
Host: That makes sense. If only one partner is talking to the AI, the AI only gets one side of the story. How did the researchers approach studying this dynamic?
Expert: The study used an interpretative qualitative research design. The researchers conducted 18 semi-structured interviews in 2025. They triangulated three perspectives: GenAI experts like developers to understand technical design, relationship psychologists to contextualize the therapeutic aspects, and actual end users who had used ChatGPT to seek relationship advice. The interviews were analyzed using summarizing content analysis.
Host: What were the key findings from these interviews?
Expert: The qualitative evidence showed that GenAI does help users self-clarify their emotions, organize their thoughts, and prepare communication strategies before they talk to their partners. For example, it can help translate emotionally charged feelings into neutral, constructive language.
Host: But there must be significant risks as well, given the one-sided nature of the interaction.
Expert: Exactly. The qualitative analysis supported that using GenAI for relationship advice carries risks of automation bias and confirmation bias. Since the AI relies entirely on what the user inputs, it tends to validate the user's one-sided and potentially biased framing of the conflict. If a user presents their partner in a negative light, the system's responses usually match that negative tone rather than offering a balanced assessment.
Host: Did the study find that users take these AI responses too literally?
Expert: Yes, the qualitative evidence supported that users often attribute unintended authority to the AI. They sometimes misinterpret its highly structured, polished responses as actual clinical diagnoses, like labeling a partner with a psychological disorder based on the AI's output. The analysis concluded that while GenAI is a useful preparatory tool, it cannot replace professional therapy because it lacks emotional intuition, nonverbal communication perception, and the ability to mediate between two partners.
Host: To keep these findings in context, what are the limitations of this study?
Expert: Because this was an exploratory qualitative study, no statistical hypotheses were tested. The findings are limited by a small sample size of 18 interviewees, retrospective self-reports, a German-speaking context, and the homogeneous use of ChatGPT among the participants.
Host: This brings us to the most crucial part for our audience of business and technology professionals. What are the practical takeaways here, especially for those designing and developing AI systems?
Expert: For technology practitioners and system designers, the study offers several concrete design implications for AI tools used in emotionally sensitive contexts. First, AI systems should explicitly communicate to users that their responses are based solely on the provided input and only reflect one perspective. Second, developers can design interface cues or reflective prompts that actively counter confirmation bias. For example, the system could prompt the user to describe the situation from their partner's point of view.
Host: What about the risk of users treating AI outputs as medical diagnoses?
Expert: To mitigate that, systems should be designed to signal uncertainty and avoid diagnostic language entirely. If a user is asking about high-stakes relational decisions, the system should suggest professional human counseling. There is also a broader need for user education to foster more critical and responsible use of GenAI in personal contexts, positioning it as a supplement to human interaction rather than a replacement.
Host: Thank you, Alex, for sharing these valuable insights. To summarize, while Generative AI is a helpful tool for individual reflection and communication preparation, it carries risks of reinforcing one-sided biases and lacks the empathy and dyadic capabilities of professional therapy. Developers of these systems need to build in active safeguards to ensure responsible use. And thank you to our listeners for tuning in to A.I.S. Insights - Turning IS Research into Business Action.
Expert: Thank you, Anna. It is great to be here to discuss this study. The researchers used the Human-AI Interaction framework, known as HAII-TIME, as an analytical lens. They conducted 18 semi-structured interviews in 2025 with GenAI experts, professional psychologists, and end users to capture different perspectives on AI-assisted relationship guidance.
Host: Let's start with the real-world problem this study addresses. Why are people turning to Generative AI for relationship advice in the first place?
Expert: Well, traditional counseling can be difficult to access due to waiting times, costs, scheduling, or social stigma. GenAI tools offer a low-threshold, private, and immediate alternative. However, the problem is that we know very little about the relational consequences and risks when these tools are used in emotionally sensitive contexts. Traditional therapy relies on a balanced, multi-perspective dialogue, whereas individual interactions with GenAI are highly vulnerable to one-sided framing and lack nonverbal cues like body language.
Host: That makes sense. If only one partner is talking to the AI, the AI only gets one side of the story. How did the researchers approach studying this dynamic?
Expert: The study used an interpretative qualitative research design. The researchers conducted 18 semi-structured interviews in 2025. They triangulated three perspectives: GenAI experts like developers to understand technical design, relationship psychologists to contextualize the therapeutic aspects, and actual end users who had used ChatGPT to seek relationship advice. The interviews were analyzed using summarizing content analysis.
Host: What were the key findings from these interviews?
Expert: The qualitative evidence showed that GenAI does help users self-clarify their emotions, organize their thoughts, and prepare communication strategies before they talk to their partners. For example, it can help translate emotionally charged feelings into neutral, constructive language.
Host: But there must be significant risks as well, given the one-sided nature of the interaction.
Expert: Exactly. The qualitative analysis supported that using GenAI for relationship advice carries risks of automation bias and confirmation bias. Since the AI relies entirely on what the user inputs, it tends to validate the user's one-sided and potentially biased framing of the conflict. If a user presents their partner in a negative light, the system's responses usually match that negative tone rather than offering a balanced assessment.
Host: Did the study find that users take these AI responses too literally?
Expert: Yes, the qualitative evidence supported that users often attribute unintended authority to the AI. They sometimes misinterpret its highly structured, polished responses as actual clinical diagnoses, like labeling a partner with a psychological disorder based on the AI's output. The analysis concluded that while GenAI is a useful preparatory tool, it cannot replace professional therapy because it lacks emotional intuition, nonverbal communication perception, and the ability to mediate between two partners.
Host: To keep these findings in context, what are the limitations of this study?
Expert: Because this was an exploratory qualitative study, no statistical hypotheses were tested. The findings are limited by a small sample size of 18 interviewees, retrospective self-reports, a German-speaking context, and the homogeneous use of ChatGPT among the participants.
Host: This brings us to the most crucial part for our audience of business and technology professionals. What are the practical takeaways here, especially for those designing and developing AI systems?
Expert: For technology practitioners and system designers, the study offers several concrete design implications for AI tools used in emotionally sensitive contexts. First, AI systems should explicitly communicate to users that their responses are based solely on the provided input and only reflect one perspective. Second, developers can design interface cues or reflective prompts that actively counter confirmation bias. For example, the system could prompt the user to describe the situation from their partner's point of view.
Host: What about the risk of users treating AI outputs as medical diagnoses?
Expert: To mitigate that, systems should be designed to signal uncertainty and avoid diagnostic language entirely. If a user is asking about high-stakes relational decisions, the system should suggest professional human counseling. There is also a broader need for user education to foster more critical and responsible use of GenAI in personal contexts, positioning it as a supplement to human interaction rather than a replacement.
Host: Thank you, Alex, for sharing these valuable insights. To summarize, while Generative AI is a helpful tool for individual reflection and communication preparation, it carries risks of reinforcing one-sided biases and lacks the empathy and dyadic capabilities of professional therapy. Developers of these systems need to build in active safeguards to ensure responsible use. And thank you to our listeners for tuning in to A.I.S. Insights - Turning IS Research into Business Action.