ChatGPT 13 times more likely than Google to go negative near the point of purchase, influencing buyer decision

BrightEdge, the global leader in enterprise SEO and AI-driven digital performance, released new data showing that AI search engines are actively evaluating brands, with each engine behaving differently. Google’s AI Overview is 44% more likely than ChatGPT to surface negative brand sentiment overall, but ChatGPT concentrates its criticism 13 times more heavily near the point of purchase. For CMOs, the result is a new form of brand risk that cannot be managed by measuring AI visibility alone.

Powered by BrightEdge AI Catalyst™, the findings arrive as over three billion people now interact monthly with Google AI Overviews and ChatGPT, roughly one-third of the world’s population. Consumers increasingly use AI not just for answers but also for brand evaluation, and AI now delivers its own editorial opinion directly in its responses.

Both engines summarize a brand’s entire digital history, including reviews, forum discussions, news coverage, and past controversies, but frame their answers and talk about brands in fundamentally different ways.

Marketing Technology News: MarTech Interview with Haley Trost, Group Product Marketing Manager @ Braze

Key Findings

1.   Negative Sentiment Is Rare, but It Reaches Millions Monthly: Google AI Overviews surface negative sentiment in approximately 2.3% of brand mentions, while ChatGPT surfaces it in approximately 1.6% of mentions.

Across billions of searches, these negative rates translate to millions of brand-negative exposures per month. Unlike a buried review on page two of search results, a negative AI response is served repeatedly to every user asking a similar question, systematically influencing demand at scale.

2.   Google Is 44% More Likely to Criticize Brands Than ChatGPT: Google and ChatGPT do not evaluate brands the same way, and the two engines are triggered by fundamentally different factors:

1. 1. Google AI Overviews skews heavily toward controversy-driven negativity, including lawsuits, boycotts, data breaches, regulatory actions, and product recalls.
   2. ChatGPT skews toward product-evaluation negativity, including compatibility limitations, feature shortcomings, and “is it worth it?” assessments.

For example, a major retailer might face negative sentiment in Google AI Overviews because of a lawsuit in the news, while in ChatGPT, the same retailer faces criticism over a specific product limitation or payment policy. Same brand, different engine, different intervention required.

Each engine draws from different source ecosystems. Google’s AI Overviews lean heavily into news-driven sourcing and controversy indexing. ChatGPT more frequently reflects product reviews, forums, and social discussions such as Reddit.

3.   ChatGPT Is 13 Times More Likely to Go Negative Near Purchase: Eighty-five percent of Google’s negative sentiment appears during informational queries, the research and discovery stage, where opinions form, and shortlists are built.

By contrast, while 68.5% of ChatGPT’s negative sentiment also appears at the informational stage, 19.4% surfaces during the consideration-to-purchase phase, 13 times higher than Google’s 1.5%. Google’s negativity gates the top of the funnel. ChatGPT’s negativity kills conversions near the point of purchase.

4.   The Engines Disagree on Which Brand to Criticize 73% of the Time: When BrightEdge analyzed overlapping prompts where both engines surfaced negative brand sentiment, Google and ChatGPT flagged different brands 73% of the time, despite responding to identical queries.

The divergence is driven by different source ecosystems: Google leans into news-driven sourcing and controversy indexing, while ChatGPT more frequently reflects product reviews, forums, and social discussions. Monitoring a single AI platform provides only a partial risk profile.

5.   Risk Profiles Vary by Industry: In electronics, both engines show elevated negativity, with Google leading because of product recalls and tech controversies. In education, Google is nearly twice as negative as ChatGPT, driven by institutional and political scrutiny.

In apparel, the pattern reverses: ChatGPT is three times more negative than Google, because fewer controversy triggers shift the dominant negativity to product-evaluation queries. A brand monitoring only one engine would miss the dynamics specific to its vertical entirely.

“For better or worse, AI is your brand’s new editorialist,” said Jim Yu, founder and CEO of BrightEdge. “Each engine characterizes your brand differently, and CMOs must treat them as distinct, dynamic environments.”

Marketing Technology News: Cross-Department Collaboration with Marketing Workflow Automation: Enhancing Alignment Between Sales, Customer Service, and Marketing Teams

What This Means for the CMO
AI is not simply indexing information. It is interpreting it, compressing a brand’s entire digital footprint into a single authoritative response. Each AI engine requires its own monitoring framework, optimization strategy, and reputation management approach. Understanding how AI talks about your brand, and measuring share of voice across search and AI, is becoming a critical metric for CMOs seeking to focus their teams and resources where it matters most.

“Sentiment monitoring across all AI engines is no longer optional,” Yu added. “It’s a revenue imperative. The brands that get ahead of this first will hold the competitive advantage.”

Buried in the Backpages No Longer

AI engines compress a brand’s historical digital footprint into a single response. Content that previously required deep navigation, including discussions from years prior, is summarized instantly. For example:

• A nearly decade-old product safety recall for a particular cell phone still appears in AI-generated responses when users search for “best phone for battery life.”
• A prompt about a major brand’s partnership with a celebrity from a years-old Reddit thread as a primary source, presenting community sentiment as established fact.
• When comparing insurance providers in California, ChatGPT mentions brands that were criticized for not renewing homeowner policies in that state a year ago.

In traditional search, these signals might have required scrolling to page two or beyond. In AI, they appear directly in the answer.

AI is not simply indexing information. It is interpreting it — and presenting that interpretation as authoritative guidance. To access the full research findings, reporters and analysts can visit the BrightEdge website. BrightEdge executives are available for briefings and interviews upon request.