AI-enabled virtual assistants (AIVAs) have become increasingly popular (e.g., Amazon Alexa, Google Home) and assist consumers with various tasks, including home automation, access to media, entertainment, and shopping. Essay 1 focuses on the outcomes of consumers' lost autonomy after information search using AIVAs versus an online search engine (e.g., Google). Drawing on research in advances in AI technology, I predict that interacting with AIVAs (versus online search engines) will lead to several consumer outcomes: decreased cognitive task performance, word of mouth (WOM) intentions, and the desire for an unrelated subsequent search. I find support for my predictions across five studies, using different tasks to assess performance (verbal and quantitative), after interactions with both real (Amazon Alexa) and fake (Halo) AIVA brands, across different respondent populations (CloudResearch, MTurk, Prolific), thereby enhancing confidence in my findings. In Essay 2, I consider a different consumer outcome - embarrassment, and also a different underlying process variable – social presence. I predict that when consumers engage in information search using an AIVA, they will subsequently experience greater embarrassment when asked about embarrassing products (e.g., condoms, medication for gas, etc.). The increased embarrassment occurs even when the information search is unrelated to the embarrassing products (e.g., searching for information on the local weather increases embarrassment related to anti-gas medication), suggesting that it is the process of interacting with the AIVA, rather than the content of such interaction which underlies this effect.
