Bospar Survey: Not a Lot of Intelligence About Artificial Intelligence

New data exposes a critical literacy gap: Public misunderstands the GEO mechanics dictating corporate visibility

SAN FRANCISCO--(BUSINESS WIRE)--Bospar, the "politely pushy" PR and marketing firm that puts tech and pharma companies on the map, today announced the results of a new nationwide survey measuring public understanding of artificial intelligence.

There’s not a lot of intelligence about artificial intelligence.

“The findings expose a serious disconnect for brands,” said Curtis Sparrer, a principal of Bospar. “While frontier and foundational models rely heavily on earned media and third-party validation to synthesize answers, the vast majority of Americans have no idea how this process works.”

Bospar tasked AI’s frontier and foundational models to draft the survey questions to determine Americans’ understanding of the fundamentals of brand perception: How AI systems surface brand information, the mechanics of Generative Engine Optimization (GEO), AI hallucinations, newsroom AI adoption, the economic concentration risk of LLM dependency and the strategic role of earned media in AI-generated answers. Then Bospar had the AI engines, including ChatGPT, Claude, Gemini, Copilot, DeepSeek, Meta and Mistral, check each other’s work to improve the survey.

Over 1,000 adults representative of the United States population answered the questions.

The Massive GEO Misconception

The public has not yet internalized that GEO is the new foundation of brand discoverability.

• Visibility blind spot: Only 29% of respondents understand that "consistent coverage in high-authority media" is the primary factor influencing a company's visibility in AI answers.
• Reputation in the AI age: Only 29% realize that the primary long-term impact of AI on PR is completely reshaping how brands build credibility online, with the majority failing to see this structural shift in how reputation is computed.
• Strategic fail: Just 41% of respondents correctly identified earning authoritative coverage in credible publications as the most effective long-term strategy to improve AI search visibility.
• Footprint flaw: Less than half (46%) understand that building a footprint across authoritative publications helps a brand become more visible over time.

The Earned Media Information Gap

As traditional search engines decline, brand visibility depends entirely on appearing in the trusted datasets that AI models scrape. Yet, the survey reveals deeply inconsistent beliefs about how AI gathers information.

• Sourcing ignorance: A staggering 72% of people believe AI visibility is driven by factors other than PR, completely missing that AI systems trust "unbiased journalism and third-party references" most when synthesizing brand information.
• Platform inconsistency: While 51% recognize LinkedIn as a key database for verified corporate identities, only 35% correctly identified that online forums are crawled based strictly on contextual relevance.
• The bright spot: The public has a clearer view of human journalists than AI systems. 58% correctly recognized that newsrooms primarily use AI for research, summarization and drafting with strict human oversight.

The Hallucination Blind Spot

The survey also found that the public is dangerously overconfident in their understanding of AI accuracy, confirming what many brands and individuals have already experienced: Uncorrected misinformation can circulate in AI-generated content without the audience recognizing it as an AI failure.

• The reality: When an AI model states a factual error, it is most often a "predicted but incorrect sequence" (hallucination).
• The public perception: Only 23% of respondents correctly identified this cause. 77% failed, attributing errors to bad data sources (26%), a coding error (30%) or a filtering rule, rather than the fundamental nature of LLMs.

Lack of Awareness of the Risk of LLM Concentration

As enterprise AI dependency deepens, the awareness gap regarding LLM concentration may leave organizations dangerously exposed to single-vendor concentration risk.

• Lack of knowledge about Microsoft CEO warning: The majority of respondents (58%) didn’t recall Microsoft CEO Satya Nadella’s warning in a June 14th essay that AI dominance will “hollow out entire industries" by allowing a handful of frontier models to absorb a sector's unique knowledge.
• Infrastructure exposure: Only 43% correctly identified loss of institutional knowledge and unique value creation as the primary risk of outsourcing AI infrastructure to a handful of LLM providers.

“This data is a brutal wake-up call for communications professionals and brand executives alike,” said Sparrer. “When nearly three-quarters of the American public can’t correctly identify how brands earn AI visibility, it tells us that the entire conversation around AI and marketing has been happening at the wrong altitude. Investing in authoritative earned media is essential to your brand’s visibility. AI doesn’t index your ad spend or your owned content. It indexes trust. And trust, in the age of generative AI, is built one credible byline, one media mention, one third-party validation at a time.”

Methodology

An online survey of 1,021 U.S. consumers, ages 18 and over, was conducted by Propeller Insights on behalf of Bospar between June 16 and June 19, 2026. Respondents opted into an online database; from there, they were targeted based on demographics. To further confirm qualifications, respondents were asked to verify their information in the survey itself, self-identifying qualifications, with the maximum margin of sampling error for the overall sampling being +/- 3 percentage points with a 95% level of confidence.

About Propeller Insights

Propeller Insights is a full-service market research firm based in Los Angeles. The firm specializes in quantitative online research with targeted consumer and B2B audiences. For more information, visit www.propellerinsights.com.