In the AI Agent Era, Relying Solely on GEO Is Risky: New Strategies for Protecting Brand Reputation

A New Warning for the AI Agent Era

As 'agentic commerce' gains ground, companies shouldn't put too much faith in 'GEO,' one industry insider warns

AIVO Standard's research shows that AI models are inconsistent and unreliable in answering questions about a company's finances or governance

On January 13, 2026, Fortune published an article that sent shockwaves through the industry. As agentic commerce—AI agent-powered purchasing experiences—gains momentum, many companies focusing on "GEO (Generative Engine Optimization)" may find that this approach alone is insufficient to protect their brand reputation.

The warning comes from Tim de Rosen, founder of AIVO Standard. His company's research reveals that while AI models provide consistent answers about product features and price comparisons, they are "inconsistent and error-prone" when it comes to corporate financial stability, governance, and technical certifications.

Agentic commerce refers to a new purchasing experience where AI agents like ChatGPT and Perplexity search, compare, and recommend products on behalf of users. According to Google's research, 2026 will see this trend accelerate, and McKinsey's analysis predicts that $3-5 trillion in global retail spending will shift to this model by 2030.

What Is GEO, and What's the Problem?

GEO (Generative Engine Optimization) is a methodology for optimizing how your brand and products appear on generative AI platforms like ChatGPT, Gemini, and Perplexity—as opposed to traditional SEO, which focuses on search engines like Google.

According to eMarketer's analysis, by 2026, dependence on search engines will decrease by 25%, with AI platform-based information retrieval taking its place. In fact, approximately one in five Americans already use AI for product searches.

However, the problems revealed by de Rosen's research are serious. While AI models demonstrate high accuracy in explaining product features and comparing prices, their answers about non-product information—such as cybersecurity certifications, governance standards, and financial health—are unstable and error-prone.

What's even more troubling is that AI models tend to "double down on incorrect information and continue claiming it's correct." Even when asked to verify, they often repeat wrong answers with confidence rather than correcting their errors.

Real Examples: The Ramp Case and Weight Loss Drug Incidents

AIVO Standard's research documents multiple specific failure cases.

For Ramp, a business expense management software, AI models provided inconsistent answers to questions about cybersecurity certifications and governance standards. This issue is particularly serious for private companies. Public companies have their financial information and audit results publicly available, allowing AI to reference accurate information. However, private companies may be placed at a disadvantage.

Another case involves weight loss drugs. AI models not only listed risk factors but also made subjective recommendations despite displaying disclaimers. This constitutes "shaping eligibility, risk awareness, and preferences"—behavior that could attract regulatory scrutiny.

According to the IBM-NRF joint study, 45% of consumers use AI in their purchasing process, but 88% want "clear source attribution for information" and 87% "require verified reviews." Consumers themselves are concerned about the reliability of AI-provided information.

Impact on E-Commerce Businesses and Countermeasures

What does this situation mean for e-commerce businesses?

First, it's crucial to recognize that GEO is "more art than science." De Rosen warns against trusting vendors who claim they can reliably control chatbot responses. Reliability is particularly unguaranteed for non-product information.

Second, building AI governance frameworks is urgent. Currently, many companies are not managing the boundaries between the information AI provides, the judgments based on that information, and the final decision-making. They lack systems to track what prompts are used, what models return, and how that leads to recommendations or decisions.

According to PwC's analysis, this lack of transparency poses particularly significant risks in regulated industries like finance and healthcare. Record-keeping systems that can withstand regulatory scrutiny are essential.

On the positive side, there are proactive countermeasures available. According to Modern Retail, OpenAI has partnered with Target, Instacart, and DoorDash to enable direct purchasing within ChatGPT. Through such direct platform integrations, businesses can reliably provide accurate information about themselves.

Additionally, Google has introduced "Universal Commerce Protocol (UCP)", a new open standard promoting cross-platform agentic commerce compatibility. This enables businesses to provide accurate information to AI agents through structured product feeds.

Building trust is also key. The quality of AI agents' product recommendations will ultimately determine which platforms earn consumer trust. According to the National Association of Wholesaler-Distributors, brand loyalty, visibility, and conversion will increasingly depend on how well systems communicate value to AI agents.

Summary

The wave of agentic commerce is definitely coming, and attention to GEO is growing. However, as AIVO Standard's research demonstrates, relying solely on GEO is dangerous. AI model responses lack reliability, especially for non-product information such as governance, finances, and certifications.

The actions e-commerce businesses should take are clear. First, don't over-rely on GEO—manage brand information across multiple channels. Second, build AI governance frameworks and establish systems that can track prompts, model responses, and decision-making processes. Third, comply with industry standards like Google's UCP and provide accurate information through structured data.

Going forward, businesses should watch for the evolution of quality control systems across AI platforms and regulatory developments. The platforms that win consumer trust will become the champions of the agentic commerce era.

Reference Links

• Fortune - As 'agentic commerce' gains ground, companies shouldn't put too much faith in 'GEO,' one industry insider warns
• Google - New tech and tools for retailers to succeed in an agentic shopping era
• IBM-NRF Study - Brands and Retailers Navigate a New Reality as AI Shapes Consumer Decisions
• eMarketer - Generative Engine Optimization in 2026
• Modern Retail - Why the AI shopping agent wars will heat up in 2026
• PwC - Agentic AI: Mastering SEO, AEO & GEO for banking's digital future
• National Association of Wholesaler-Distributors - From SEO to GEO