OpenAI has begun testing advertising features for free and low-cost subscription users to alleviate the high operational costs. This move has sparked strong opposition from users, being criticized for harming user experience and trust. Competitor Anthropic seized the opportunity to mock OpenAI, prompting a rare fierce response from OpenAI's CEO. The underlying motivation is to support its hundred-billion-dollar financing negotiations by demonstrating sustainable profitability to capital markets, potentially accelerating the commercialization process of the entire AI industry.
OpenAI officially announced on Monday that it has started testing advertising features among certain user groups of ChatGPT in the United States, marking a decisive step in its path toward commercial monetization. Although OpenAI emphasized that this initiative aims to cover expensive infrastructure costs through advertising revenue to maintain free services, the decision immediately triggered widespread controversy on social media and strong opposition from users.
This test is only available to logged-in users of the free version and the newly launched low-cost subscription service 'Go' (USD 8 per month), while advanced subscription users such as Plus, Pro, Team, and Enterprise remain unaffected. OpenAI has committed to visually distinguishing advertisements from organic responses generated by the chatbot. Advertisers will not have access to specific user conversation content. The company is attempting to strike a balance between expanding revenue sources and maintaining user trust but still faces severe challenges regarding privacy protection and user experience.
This commercial move has rapidly intensified competition within the AI sector. Major competitor Anthropic capitalized on the moment by airing an advertisement during the Super Bowl satirizing competitors’ ad placements, implying that commercial ads would undermine the utility of AI assistants. In response, OpenAI’s CEO made a rare, aggressive rebuttal, accusing the rival of being 'dishonest.' Despite the noisy public discourse, market analysts generally believe that OpenAI urgently needs to establish a sustainable revenue model to support its ongoing hundred-billion-dollar financing negotiations.
Clear Path to Commercialization
According to OpenAI’s official announcement, the core logic behind this advertising test lies in 'task alignment.' ChatGPT currently boasts hundreds of millions of users, and maintaining services for the free and Go tiers requires substantial investment in computing power and infrastructure. OpenAI stated that introducing an advertising model helps provide high-quality AI services to a broader audience while supporting continuous iteration in intelligence and capabilities.
In terms of execution, OpenAI adopted a matching mechanism similar to traditional search engines. The system determines displayed content based on the user’s conversation topics, past chat history, and ad interaction records. For example, when a user searches for recipes, the interface may display ads for food delivery or meal kits. OpenAI emphasized that all advertisements will be clearly labeled as 'sponsored content,' and advertisers cannot influence the answers generated by ChatGPT, ensuring the independence of responses.
For users who do not wish to see advertisements, OpenAI offers the option to upgrade to premium paid plans. Additionally, free-tier users can manage ad personalization or opt out of ads in their settings, but in exchange, their daily quota of free messages will be reduced.
Privacy Commitment and User Control
Addressing external concerns about potential privacy breaches, OpenAI elaborated on its data protection mechanisms in the announcement. The company explicitly stated that advertisers cannot access users’ chat history, memory functions, or personal details. Advertisers will only receive aggregated data about ad performance, such as views and clicks.
OpenAI has committed that during the testing period, the system will automatically block advertisements targeting users under the age of 18. Additionally, it prohibits displaying ads near sensitive and regulated topics such as health, mental health, or politics. Users retain a certain level of control on the client side, including rejecting specific ads, providing feedback, deleting ad data with one click, and managing personalized ad settings at any time.
Although OpenAI claims that model training is isolated from ad content, and unless users actively click and request an explanation of the ad, the model itself is “unaware” of the ad content, whether this mechanism can completely eliminate user concerns remains to be seen.
Market Response and Competitive Dynamics
OpenAI’s move has sparked intense negative comments on social media. According to user feedback on the X platform (formerly Twitter), many users expressed disappointment with “AdsGPT.” User Nicole D pointed out that within four days, OpenAI seemed to be abandoning tools crucial for the neurodivergent community and questioned the company’s so-called “protection of trust.” Another user, David Stark, warned that this action removes the last remaining advantage of ChatGPT and could lead to user attrition. Some users even stated bluntly that if ads are added, the product would no longer qualify as a true intelligent assistant.
Competitor Anthropic keenly captured this sentiment. According to TechCrunch, in its Super Bowl commercial, Anthropic mocked OpenAI’s strategy by showcasing a vacant-eyed AI chatbot alongside poor ad placements. This marketing offensive directly infuriated Altman, who publicly criticized Anthropic as an “authoritarian company.”
Despite facing resistance from some users, there were also voices of understanding. MomentumAI noted on social media that as long as ads do not affect the generated prompt responses, the presence of ads in free products is acceptable.
Financing Prospects and Product Iteration
While launching the ad testing, OpenAI is also accelerating product iteration and capital operations. Media reports indicate that OpenAI’s recently released new programming model, GPT-5.3-Codex, has received positive market feedback, with user growth increasing by approximately 50% and monthly growth recovering to over 10%.
More critically, OpenAI is advancing a large-scale financing plan. Reports suggest that the total amount of this round of financing is expected to reach $100 billion, divided into two parts: the first portion of approximately $50 billion comes from Microsoft, NVIDIA, and Amazon, with follow-up investors like SoftBank expected to contribute an additional $30 billion.
Analysts point out that OpenAI’s rush to test the advertising model aims both to alleviate operational cost pressures and to demonstrate the scalability of its business model to the capital markets. Once the “AI + advertising” commercial loop is established, it will not only provide OpenAI with self-sustaining capabilities but may also trigger a chain reaction in the large model industry, prompting other manufacturers on the sidelines to accelerate their commercialization processes.
Editor/Doris
