The AI Trust Paradox: Americans Embrace Chatbots But Don't Believe Them

MIAMI, FL – November 24, 2025 – Artificial intelligence has seamlessly woven itself into the fabric of American life, but a new study reveals a profound paradox at the heart of this digital transformation. While millions now rely on AI chatbots for daily tasks, a deep-seated skepticism is forcing users to adopt a “trust but verify” approach, signaling a critical maturation point for the AI industry.

A new survey from ChatOn, an AI application from developer AIBY with a claimed user base of over 85 million, finds that chatbots have become indispensable tools. A staggering 74% of Americans use them to find information, 65% for writing and editing communications, and 54% for brainstorming. Yet, this widespread adoption is shadowed by a significant trust deficit: 39% of those same users feel compelled to verify AI-generated information using Google or other sources.

This gap between utility and reliability highlights a crucial challenge for the industry. As AI moves from novelty to infrastructure, the focus is shifting from what the technology can do to how well it can be trusted. The data suggests that for a large segment of the population, the answer is: not completely.

The Rise of the Skeptical User

The ChatOn survey paints a clear picture of a user base that is increasingly savvy and cautious. The instinct to fact-check is not an isolated habit but part of a broader behavioral shift driven by direct experience with AI’s shortcomings. Users report frequently encountering AI “hallucinations,” with 39% stating they “sometimes” receive irrelevant responses, 36% finding outdated information, and 33% noticing outright contradictions in the answers provided. These findings are consistent with external studies, such as a February 2025 analysis from UX Tigers which found that even advanced models like GPT-4 can hallucinate nearly 30% of their cited references.

This awareness of AI's fallibility directly fuels user caution, particularly regarding data privacy. According to the survey, 54% of respondents actively avoid sharing sensitive personal information with chatbots, 42% refrain from uploading confidential files, and 36% steer clear of discussing work-related data. These concerns are not unfounded and reflect a wider market sentiment. A late 2024 survey from Deloitte’s “Connected Consumer” series revealed that 79% of consumers believe tech providers are not transparent about their data policies.

This climate of distrust is compounded by academic research, including a recent Stanford study which concluded that the privacy policies of major chatbot developers are often vague, allowing for long data retention periods and the use of private conversations for model training. For businesses integrating these tools into their workflows, this user skepticism presents a tangible operational risk, forcing them to establish strict governance protocols to prevent the inadvertent leakage of proprietary information.

A New Digital Divide: The Emergence of AI Literacy

Beyond simple usage, the survey reveals the emergence of a new form of digital divide, one not based on access but on proficiency. Nearly half of users (49%) rate their skills as 'intermediate,' with another 24% identifying as 'advanced.' These groups are not passive consumers; they are actively developing a sophisticated skill set that could be defined as AI literacy.

This new literacy is characterized by specific habits aimed at overcoming the technology’s limitations. The most common techniques include asking follow-up questions to refine answers (48%), rephrasing prompts to improve results (42%), and experimenting with different prompts to test the AI’s boundaries (46%). These behaviors separate effective AI users from casual ones, who may be more susceptible to misinformation or less capable of extracting real value from the tools.

Interestingly, other industry research supports the idea that experience breeds skepticism. A September 2025 study by Rev found that daily AI users are 14 times more likely than casual users to double-check the AI's work, reinforcing the notion that proficiency and verification go hand-in-hand. This is the user base that understands the technology’s power but is also acutely aware of its flaws. As Dmitry Khritankov, Product Director at ChatOn, noted in the release, “familiarity doesn't equal mastery.” This gap presents a clear mandate for the industry: to build tools that are not only powerful but also transparent, reliable, and safe.

Navigating a Market of Giants

ChatOn's decision to publish this survey can be seen as a strategic move in a fiercely competitive market. The AI chatbot space is dominated by titans like OpenAI’s ChatGPT, which commands over 80% of web traffic in the category, alongside formidable offerings from Google, Microsoft, and Anthropic. For a smaller, app-focused player like AIBY’s ChatOn, competing on model size or brand recognition alone is an uphill battle.

Instead, ChatOn appears to be positioning itself by addressing the very user fears its survey highlights. The application differentiates itself by functioning as a hub that integrates multiple leading large language models—including GPT-5, Gemini 2.5 Pro, and Claude 3.7—theoretically allowing users to access the best tool for a given task. Furthermore, it bundles a suite of practical features like a “Document Master” for file analysis, a “Web Analyzer” for up-to-date information, and an integrated AI image generator.

By focusing on these user-centric features and publicly acknowledging the market's concerns around trust and safety, companies like AIBY are signaling a strategic pivot. The race is no longer just about launching the most powerful model; it is about building the most reliable and complete user experience. This approach, which prioritizes transparency and directly tackles the pain points of accuracy and privacy, may offer a viable path for differentiation. As the market matures, the companies that win deeper user loyalty may not be the ones that shout the loudest, but the ones that listen most closely to an increasingly intelligent and discerning customer base.