When they arrived on the scene, voice assistants seemed like sci-fi made real.
Once hailed as revolutionary, they are now seeing a downshift in consumer enthusiasm. The share of U.S. consumers using voice assistants has sat at 60% this year. This ambivalence is evident across generations. Declining trust and satisfaction with voice technology appears to be a key factor. This suggests that consumers’ expectations for seamless and reliable performance have not been met. But the performance improvements that deeper GenAI integration promise could reverse declining adoption rates for voice assistants across generations.
These are some of the findings explored in “GenAI and Voice Assistants: Adoption and Trust Across Generations.” This PYMNTS Intelligence special report examined consumers’ preferences and behaviors surrounding voice assistant and GenAI technologies. It drew on a census-balanced survey of 2,721 U.S. consumers that asked about their use of and interest in AI-enabled technologies conducted between June 5 and June 21.
Trust Fall: Declining Confidence in Voice Technology
Trust in voice assistants has declined. Fewer consumers surveyed this year said these technologies will become as smart and reliable as humans when compared to last year’s survey. In March 2023, 73% of consumers expressed confidence in the future capabilities of voice assistants. By June 2024, this figure dropped to 60%. The overall decline was particularly pronounced among millennials and bridge millennials, who were initially the most enthusiastic about voice technology.
Part of the problem is, for many consumers, this technology has already outstripped much of its early promise. As a result, 40% of consumers either doubt or remain unsure about whether voice assistants will ever be as smart and reliable as humans. This figure represents an increase in consumer skepticism, jumping from 27% in March 2023.
Voice Assistants Not Much Help When Needed Most
Forty-three percent of U.S. consumers would trust a humanlike voice assistant to call for help in an auto accident. This figure is down from 50% last year and is mirrored across various moments of unexpected need. One such scenario is notifying family when an appointment runs late and that someone needs to pick up the kids. The share of millennials who would trust voice assistants in this scenario is now 48%. For zillennials — a microgeneration situated between Generation Z and millennials — the share is 49%. Both groups reported trust levels in this scenario above 55% last year.
Likewise, Generation X consumers’ trust in voice assistants in this scenario dropped to 34% earlier this year. Baby boomers and seniors showed a similar decline, falling from 33% last year to 26% in June. Both groups exhibited the lowest trust levels in using voice assistants in more serious situations across generations as well. Forty-three percent of Gen X and 35% of baby boomers and seniors would trust voice assistants after an auto accident.
The primary reasons for the decline in trust may be a perceived lack of improvement and reliability in voice assistants. Consumers, especially younger ones, often have high hopes and lofty expectations for technological advancements. Early adopters and digital natives might even tolerate some flaws and inconsistencies as emerging technology matures. However, patience will wear thin when these issues persist over time, leading to decreased trust at scale.
Generational Differences in the Adoption and Integration of GenAI
GenAI could upend the status quo by improving the accuracy and precision of voice-enabled technologies. While voice assistants struggle to maintain their user base, GenAI has seen rapid and robust adoption across all generations. Generational trends showed that Gen Z led in familiarity, at 82%, followed by zillennials, at 78%. At the same time, familiarity among baby boomers and seniors jumped from 23% to 41%.
Usage patterns reflected this growing familiarity. In the past 90 days, 34% of U.S. consumers used GenAI. Gen Z showed the highest usage, at 63%. Zillennials also demonstrated engagement, with 20% using GenAI at least weekly. Frequent use highlighted the role of GenAI in fulfilling quick information retrieval, as 47% of consumers used GenAI for this purpose. In addition, 41% said they found it very useful for editing text.
GenAI Could Reintroduce Smartphone Users to Their Voice Assistants
If GenAI does indeed ride to voice assistants’ aid, its journey must start with smartphones. Gen Z led in using voice assistants built into smartphones, with a 42% usage rate this year. Zillennials (41%) and millennials (38%) followed. On the other hand, millennials used voice-activated devices more often than other consumers. Millennials and bridge millennials, at 32% each, used voice-activated devices more than any other consumer segment. Baby boomers and seniors showed the lowest usage rates across categories, particularly when using voice capabilities in cars and wearables.
Conclusion
The perceived utility of GenAI has remained strong among all consumers this year. The overall share of U.S. consumers who think GenAI is useful for information retrieval remained stable. It was 62% last year and 63% as of June. This stability suggests that while trust in voice technology wanes, confidence in GenAI’s capabilities remains strong.
Generational differences highlighted distinct adoption and integration patterns. Zillennials, as digital natives, are more comfortable navigating new technologies. Zillennials’ willingness to integrate these tools into daily tasks contrasts sharply with the more cautious approach of older generations. However, the waning trust in voice assistants should serve as a cautionary tale.
Methodology
“GenAI and Voice Assistants: Adoption and Trust Across Generations,” a PYMNTS Intelligence exclusive report, was based on a survey conducted from June 5 to June 21 that examined the use of GenAI and voice assistants across generations. The study drew insights from this census-balanced panel of 2,721 U.S. consumers, which included an oversampling of 1,201 consumers born between 1991 and 1999 to facilitate a deeper analysis of the zillennial microgeneration. Population weights were used to ensure analysis remained representative of the U.S. adult population.
