Vivek Shah Highlights the Shift from Journalistic to Marketing Sources in AI Responses

Vivek Shah, the CEO of Ziff Davis-the parent company of ZDNET-has issued a strong advisory for consumers to exercise caution when using AI chatbots for guidance on significant purchases. While acknowledging the revolutionary potential of AI technologies, Shah warns that many AI-driven platforms increasingly rely on marketing content rather than impartial journalism, which can skew the quality and reliability of the information provided.

“The origin of information is crucial,” Shah emphasized. “When you examine the citations in large language model (LLM) chatbots, you’ll notice a growing trend where references have shifted from credible journalistic outlets to predominantly marketing sources.”

Evaluating AI Chatbots: The Importance of Source Transparency

Shah’s concerns extend beyond ChatGPT to other AI tools such as Google Gemini, Perplexity, and Anthropic’s Claude. Although these platforms often include citations within their conversational responses, the visibility and accessibility of these sources vary significantly. Some make it difficult for users to verify the origins of the information, potentially leading to less informed decisions.

He encourages consumers to actively investigate the references behind AI-generated advice, asking themselves whether they would have trusted those sources before the advent of AI chatbots. This critical approach is essential, especially when making high-stakes purchases.

Comparing AI Chatbots: A Case Study on Product Recommendations

To illustrate the disparity in source quality, a recent test was conducted by querying four popular AI chatbots about the purchase-worthiness of the Meta Ray-Ban Display glasses. The results varied: Claude and Gemini predominantly cited vendor-related sources, while Perplexity and ChatGPT leaned more on independent publishers. Notably, Perplexity presented its sources most prominently and made them easily accessible for verification.

It’s important to note that AI responses can differ between users and even fluctuate with repeated queries, underscoring the need for cautious interpretation.

Balancing AI Optimism with Intellectual Property Concerns

Despite his reservations, Shah maintains a positive outlook on AI’s transformative capabilities. Overseeing a portfolio of companies that specialize in impartial, third-party product advice, including ZDNET, he recognizes AI’s potential to enhance business and consumer experiences.

Shah also revealed that Ziff Davis is exploring opportunities to license its trusted data to future AI systems, aiming to improve the quality of AI-generated content. However, he remains vigilant about intellectual property rights, emphasizing that the challenge lies in ensuring AI respects content ownership while evolving.

“My concern centers on intellectual property, not on AI’s transformative impact,” Shah explained. “AI will undoubtedly reshape our lives and industries, but it must do so responsibly.”

Practical Advice for Consumers Navigating AI-Driven Shopping

As AI becomes increasingly integrated into online shopping-such as ChatGPT’s new capabilities to make purchases on behalf of users-consumers should remain discerning. Verifying the credibility of AI-sourced information and understanding the nature of cited references can help avoid biased or incomplete advice.

For example, when considering a high-value tech gadget, cross-referencing AI recommendations with established review sites or trusted expert opinions remains a prudent strategy.