OpenAI Pivots After Shopping Setback, Faces Challenge of Building a ‘Clean Brand’ Amid Enterprise Push

Weak performance in B2C expansion leads to scaling back of some new ventures
IPO preparations proceed alongside a shift toward productivity tools
Corporate ethics versus pragmatism emerges as a factor shaping brand value

OpenAI has clarified its strategic direction toward enterprise-focused AI by announcing the discontinuation of its ChatGPT-based shopping feature. The move reflects a broader effort to transform its AI models into corporate productivity tools while strengthening coding capabilities. As the company expands consulting and partnerships for enterprise clients and reorganizes internally in preparation for a potential public listing, its overall business structure is undergoing a realignment. At the same time, divergent approaches to military cooperation and ethical standards compared with competitors are bringing renewed attention to OpenAI’s brand positioning as a predictable partner.

Consumer-facing services face lukewarm response

According to industry sources, OpenAI recently decided to discontinue “Instant Checkout,” a shopping feature introduced in the second half of last year through ChatGPT. The feature, which allowed users to search for and purchase products from external online retailers within ChatGPT, was part of the company’s effort to diversify its business. When it launched in September, it drew attention with participation from major global retailers including Walmart, Shopify, and the handmade marketplace Etsy. However, six months after its release, frequent errors—such as outdated product information not being properly reflected—led to declining user engagement.

OpenAI has also expanded into other areas, including social media and healthcare. However, the actual performance of most services has fallen short of expectations. A representative example is “Sora,” a video-generation AI-based social platform. First introduced in October 2024, Sora recorded one million downloads within five days of launch. Yet just two months later, in December, downloads fell 32% month-over-month, followed by an additional 46% decline in January. In-app purchase volumes also steadily declined, and by March last year, the app had dropped out of the top 100 in the U.S. App Store’s free app rankings.

Meanwhile, the competitive landscape has evolved rapidly. Google has expanded its presence in the consumer AI market with its image-generation AI “Nanobanana,” while Meta has intensified competition by launching a similar video-generation service, “Vibes.” At the same time, restrictions on the use of intellectual property for content creation and growing user resistance to the use of facial data have imposed further constraints on service expansion. These conditions have led to the assessment that OpenAI has not secured a stable revenue model in the consumer services segment.

In response, OpenAI is restructuring its business focus around enterprise markets. Internally, the company has finalized a strategy to concentrate resources on advancing generative AI models and developing corporate productivity tools. To support this shift, it plans to expand its workforce from approximately 4,500 employees to around 8,000 within the year and is recruiting specialized personnel, including “technical ambassadors,” to assist enterprise clients in adopting OpenAI tools. These measures are interpreted as a response to its primary competitor, Anthropic, which has signaled ambitions to replace traditional software markets through its AI agent “Claude Cowork.”

Need for reputation management ahead of IPO

OpenAI’s expansion of its enterprise consulting division is closely tied to its IPO timeline. CNBC, citing sources familiar with the matter, reported that OpenAI is considering going public within the year, with a strong possibility of an IPO as early as the fourth quarter. Within the company, there is a clear shift toward directing the attention of employees and investors to enterprise-focused operations. CNBC described this as reflecting a judgment that expanding the enterprise client base is more directly linked to financial structure than continuing to scale consumer services.

In this process, the role of ChatGPT itself is being redefined. Fidji Simo, CEO of OpenAI’s applications division, stated at an internal meeting earlier this month, “Our current goal is to convert 900 million users into high-compute users,” emphasizing the need to aggressively expand enterprise support and secure high-value use cases. At the same time, financial and organizational restructuring is underway. OpenAI recently hired former Block executive Ajeet Singh and former DocuSign CFO Cynthia Gaylor. Industry observers expect Gaylor to oversee investor relations at OpenAI.

A key variable is that expanding into enterprise markets introduces new requirements. As AI adoption spreads across business operations, companies are increasingly evaluating not only productivity gains but also brand risks. Collaborating with partners involved in social controversy can raise external communication challenges, while internally, firms must assess data usage practices and accountability for outputs. In enterprise environments, factors such as security, predictability, controllability, and external reputation are becoming as critical as model accuracy. For corporate clients, whether a company maintains a “clean brand” is emerging as a decisive factor in contract evaluations.

Military partnerships put corporate ethics to the test

Recent developments surrounding Anthropic’s military cooperation with the U.S. Department of Defense illustrate this dynamic. During negotiations over the military use of its AI models, Anthropic demanded two exceptions: a ban on “mass domestic surveillance” and on the use of its technology for fully autonomous lethal weapons. The Department of Defense rejected these conditions, viewing them as constraints on operational effectiveness. Pentagon spokesperson Sean Parnell stated that “the U.S. military has no interest in unlawful domestic surveillance or autonomous weapons without human oversight,” while Deputy Secretary Emil Michael added that “we cannot seek permission from a private company to shoot down enemy drone swarms targeting Americans.”

The dispute escalated beyond policy differences into concrete administrative action. On February 27, the Department of Defense designated Anthropic as a “supply chain risk” entity, a classification typically reserved for adversarial foreign actors, making it the first instance of a U.S. company receiving such a designation. President Donald Trump also ordered all federal agencies to cease using Anthropic’s technology, citing concerns over reliability. While some political voices warned that “a government forcing the deployment of AI weapons without safeguards is itself alarming,” the Trump administration dismissed such concerns.

Industry observers have focused on what this episode reveals about Anthropic’s brand identity. Prior to being designated a supply chain risk, Claude had been the only large language model authorized to handle classified Department of Defense information. Nevertheless, Anthropic maintained guidelines restricting the use of its models in violent military operations. This stance led to conflict with the government but was also seen as consistent with the company’s positioning as a provider of “safe and ethical AI.” Such positioning is likely to influence corporate clients that weigh not only technical performance but also the brand integrity of AI providers when making adoption decisions.

Meanwhile, the gap left by Anthropic has been quickly filled by OpenAI. According to The Wall Street Journal, OpenAI signed a classified contract with the U.S. Department of Defense earlier this month and has begun a project to deploy its AI models within military networks. At an internal meeting, CEO Sam Altman stated that “while individuals may have differing views on the U.S. strikes on Iran, the company is not in a position to judge such matters,” adding that “our focus is on providing technical guidance and building safety frameworks.” The remarks are interpreted as reflecting awareness that differing stances on military cooperation could shape evaluations of corporate responsibility and standards.