AI Usage Soars Despite Trust Issues: The Paradox of Generative AI Adoption

The Rise of Generative AI in Online Search

In a remarkable shift in internet usage patterns, generative AI tools, particularly ChatGPT, have seen explosive growth since late 2022. ChatGPT now processes approximately 2.5 billion user queries daily, with 13% originating from the United States 2

2

. This surge in adoption has positioned ChatGPT as an increasingly central component of global internet interaction, evidenced by its status as the most downloaded global application in April 2025 and surpassing the combined App Store downloads of popular social media platforms like TikTok, Facebook, and Instagram in June 1

1

2

2

.

Industry Giants Respond to the AI Revolution

The rapid rise of ChatGPT has prompted swift responses from industry leaders. Google, which processes approximately 5 trillion queries annually, has pivoted to integrate generative AI-powered features, most notably AI Overviews, into its search engine 1

1

. This move represents a significant shift in the company's approach to online search.

Meanwhile, new players are entering the field with AI-centric offerings. Perplexity, an online search startup focused on generative AI, has launched its own web browser, Comet, to compete with established browsers like Chrome and Safari 1

1

. Google has also debuted AI Mode, a feature that incorporates generative AI more deeply throughout the search experience, operating similarly to Perplexity's core offerings 1

1

2

2

.

The Trust Paradox: Usage vs. Confidence

Despite the surging adoption of generative AI in online search, user trust remains a significant concern. A recent survey of over 1,100 Americans revealed a striking paradox:

• Only 8.5% of respondents said they "always trust" the information they receive from Google's AI Overviews
• Approximately 21% expressed zero trust in the feature's ability to surface accurate information
• More than 40% of respondents rarely or never click on accompanying web links provided in AI Overviews to view the source material 1

  1

  2

  2

This data paints a picture of users interacting with generative AI by default in their daily online activities, while simultaneously harboring little to no confidence in the accuracy of the information they receive.

Factors Influencing AI Trust

Research indicates that trust in AI systems is nuanced and context-dependent. A study found that the majority of people trust ChatGPT more than human experts in at least one domain. However, this trust diminishes when the chatbot discusses sensitive subjects like legal or medical advice 1

1

.

The perception of AI trustworthiness is also influenced by communication style. More sycophantic AI-generated responses tend to be perceived as less trustworthy compared to responses with a more neutral, less explicitly flattering tone 1

1

.

The Challenge of AI Hallucinations

A significant challenge facing generative AI systems is the phenomenon of "hallucination" – providing inaccurate information with the veneer of truth. While large language models are generally trained to maximize facticity, they can veer unpredictably into producing false or misleading information 1

1

2

2

.

Industry Efforts to Build Trust

Recognizing the importance of user trust, leading AI developers like OpenAI and Anthropic have launched initiatives aimed at interpreting and understanding the intricate processes taking place within AI systems. These efforts are dual-purposed: to improve system performance and to boost user confidence and trust in AI-generated information 1

1

2

2

.

As generative AI continues to reshape the landscape of online search and information retrieval, the industry faces the ongoing challenge of balancing rapid adoption with the critical need for user trust and reliability.