OpenAI Pivots to Advertising Model to Fund Rising Infrastructure Costs
San Francisco, Wednesday 21 January 2026
With only 5% of 800 million weekly users paying for subscriptions, OpenAI initiates impression-based ad trials to subsidise its massive infrastructure costs.
The Commercial Pivot
OpenAI has formally approached a “small pool” of advertisers to secure commitments for its new advertising inventory, marking a definitive transition from a purely subscription-reliant business model [1]. According to reports emerging on Tuesday, 20 January, the Microsoft-backed AI giant is seeking initial spending commitments of under $1 million per advertiser for a trial period scheduled to commence in early February [1]. Notably, the company has opted for a pricing structure based on ad views (impressions) rather than the industry-standard “cost-per-click” model, suggesting OpenAI is positioning its chatbot as a premium brand-awareness channel rather than a direct performance marketing tool [1]. While a self-service platform is currently in development, the initial phase requires direct engagement, with the company reportedly working to operationalise automated buying systems in the near future [1].
Tiered Access and User Impact
The advertising rollout will target users of the free tier as well as subscribers to the recently launched ‘ChatGPT Go’ plan, which became available on 16 January 2026 for $8 per month [2][3]. This creates a clear demarcation in the user base: those willing to pay a premium for an ad-free experience via the Plus, Pro, and Enterprise plans—with the Plus tier costing roughly €23 ($23) per month—and those on lower-tier plans who will now be monetised through display advertisements [2][4]. Reports indicate these advertisements will occupy approximately one-third of the screen, appearing at the bottom of chat responses when relevant products or services are identified [3][4].
The Economics of Scale
The impetus for this commercial expansion appears to be the sheer magnitude of OpenAI’s operational costs. Despite projecting $13 billion in revenue for the current year, the company anticipates burning through approximately $9 billion, driven largely by capital-intensive infrastructure requirements [3]. In 2025 alone, OpenAI reportedly committed to infrastructure deals valued at $1.4 trillion, a staggering sum necessary to support the computational demands of next-generation models [2]. With only 5% of its 800 million weekly users currently converting to paid subscriptions, the introduction of advertising revenue is a necessary lever to sustain these expenditures and prepare for a highly anticipated initial public offering (IPO) [1][3].
Market Dynamics and Competition
Competitive pressure is also reshaping the landscape. OpenAI’s dominance has waned significantly over the past year; its market share has contracted from 87% to 65%, while Google’s Gemini has surged, recording a 266.667 per cent increase in its share to reach 22% [3]. This erosion of market dominance highlights the urgency for OpenAI to diversify its revenue streams before its liquidity is challenged, with some analysts warning that the company’s cash reserves could be depleted by mid-2027 without significant financial intervention [3].
A Departure from Principles?
This move represents a significant ideological shift for CEO Sam Altman, who as recently as 2024 described advertisements in AI as “uniquely disturbing” and a “last resort” [3]. To mitigate privacy concerns, OpenAI has assured users that conversation data will not be sold to third parties and that advertisements will be clearly distinguished from organic search results [2]. However, the juxtaposition of ads within a conversational interface may test user patience in an increasingly fragmented market where privacy-focused competitors are beginning to gain traction [3][4].