OpenAI says ChatGPT’s web-search query volume has reached 20% of Google’s search volume, according to Adam Fry, the company’s consumer product lead for ChatGPT. The claim matters to businesses, developers and users because it suggests that conversational AI is becoming a meaningful route to online information, even though Google remains far ahead and the comparison is not fully documented.
What OpenAI says about ChatGPT search
Fry made the announcement at a press event on Monday. OpenAI’s definition is narrower than all ChatGPT activity: it counts an input when it triggers a web search, rather than requests such as defining a word or explaining a simple how-to task.
That distinction is important. The 20% figure does not mean that ChatGPT handles one-fifth of all user questions, nor does it establish that ChatGPT is close to replacing Google. It indicates that the volume of ChatGPT queries classified by OpenAI as searches is equivalent to one-fifth of the volume OpenAI is comparing against Google.
The source does not identify where the Google search-volume figure came from. It also provides no breakdown by country, device, query type or time period. As a result, the announcement supports a directional comparison, not a like-for-like independent measurement.
What is new—and what remains unconfirmed
The new information is OpenAI’s reported search-volume milestone and its explanation of what counts as a search. The material supplied does not confirm a new ChatGPT search feature, a product redesign, a rollout schedule, geographic availability or eligibility requirements.
It also does not describe how web results are selected, how citations or source checking work, or whether ChatGPT search has different privacy and security implications from conventional search. Questions about data retention, publisher traffic and the handling of sensitive business queries therefore remain unanswered by this announcement.
Implications for teams and developers
For companies that depend on online discovery, the claim is a reason to monitor conversational search separately from traditional search traffic—not proof that existing search strategies should be abandoned. Teams may need to assess whether their documentation, product pages and public support material are clear enough to be interpreted in answer-based interfaces, while continuing to measure actual referrals and conversions.
For example, a hypothetical software company could compare support requests and website visits originating from ChatGPT with those from Google before changing its content workflow. The potential advantage is earlier visibility into how customers ask multi-step questions; the limitation is that OpenAI has not supplied enough detail here to show which industries, regions or query categories drive the reported share.
For now, the practical next step is measurement rather than migration: treat OpenAI’s 20% figure as a company-reported indicator, verify impact using first-party analytics, and wait for clearer methodology and availability information before making major search or privacy decisions.
资料来源 原创报道. 本文包含独立的编辑背景和实务分析。.
更多关于人工智能的见解： 浏览我们的 最新的AI自动化与技术文章, ，或探索 即用型自动化解决方案.
