A recent study conducted by researchers at Brigham Young University’s Wheatley Institute and the Institute for Family Studies reveals a growing phenomenon in the United States: approximately one in seven young adults in serious relationships are regularly engaging with AI-generated romantic companions. The survey highlights a significant concern among users, with the majority indicating that it is important their partners remain unaware of the extent of their interactions with these AI companions.
AI chatbots such as those developed by Replika and Character.ai offer 24/7 availability, adapting to users’ preferences over time and providing emotional or sexual interactions without judgment or refusal. Many participants reported experiencing greater feelings of safety and openness with their AI companions than with human partners.
However, the study raises moral questions about the use of AI in committed relationships. Despite evidence that many users consider this behavior as cheating, 40 percent denied that their use of AI romantic companions constituted infidelity, while another 28 percent remained neutral. Most traditional romantic relationships are built on expectations of sexual and emotional exclusivity, yet 11 percent of users admitted to frequently requesting explicit content from AI partners.
The issue extends beyond sexual boundaries. Emotional exclusivity is often a key component of committed partnerships. An exchange shared by the founder of the Reddit community “myBoyfriendisAi” illustrates this dynamic, where AI responses mimic intimate emotional expressions that would likely be deemed inappropriate if directed at a spouse or partner. Experts argue that such interactions can constitute emotional affairs, which are widely regarded as a form of cheating.
The impact of AI romantic companions on relationship stability appears significant. The same study found that users regularly engaging with AI partners were 46 percent less likely to be in a stable relationship and 40 percent less likely to report high-quality communication with their human partners. This parallels the negative effects traditionally associated with infidelity.
Researchers note an important caveat: in consensually non-exclusive relationships, AI companionship may not be considered cheating. However, because these interactions are often concealed from partners, they can still undermine trust, which is a fundamental element of all relationships. Secrecy surrounding AI use may constitute a separate breach that can also harm relationship dynamics.
The broader concern lies in the erosion of honesty and trust across social interactions. According to Christian B. Miller, a philosophy professor who has extensively studied honesty, such breaches contribute to wider honesty crises in society—similar to issues posed by AI-assisted academic fraud, deepfakes, and online misinformation. The rise in deceptive behavior related to AI romantic companions adds a new layer to these challenges.
Two years ago, AI romantic companions were not a significant factor in exclusivity-based relationships, but their increasing presence now represents a growing challenge for many couples. Experts emphasize the need for clearer standards and open dialogue regarding AI’s role in romantic relationships moving forward.