The intense competition for dominance in the burgeoning enterprise artificial intelligence market is experiencing a significant shift, with new data from corporate credit card and expense management firm Ramp indicating that OpenAI is beginning to regain market share against its primary rival, Anthropic, among U.S. businesses. This development offers a crucial glimpse into the performance of these closely watched AI powerhouses, especially as both companies approach potential initial public offerings (IPOs) and their financial details remain largely private.
Ramp’s latest analysis, drawing from the spending patterns of over 70,000 American businesses that process billions through its bill pay and corporate card products, reveals a volatile yet expanding landscape. While Anthropic had overtaken OpenAI as the leader among Ramp’s paying business users in May, reaching 41% market share compared to OpenAI’s 39%, OpenAI is now demonstrating renewed momentum. As of July, Anthropic held nearly 44% of this market segment, while OpenAI had climbed to nearly 40%. More notably, Ramp economist Ara Kharazian points out that OpenAI is currently growing faster than Anthropic among this segment in Q3 to date. This rapid reversal underscores the highly dynamic nature of the enterprise AI sector, where market positions can shift dramatically within months, or even weeks.
A Deeper Look into Ramp’s Findings and Market Dynamics
Ramp’s dataset, while robust, offers a specific lens into the market. Its customer base spans various industries, but given its popularity as a Silicon Valley corporate credit card provider, it naturally skews towards the technology sector. This provides a valuable, albeit not all-encompassing, proxy for enterprise AI adoption, particularly among fast-growing and tech-forward companies. It’s important to acknowledge that this data does not represent the entire market, as it excludes very large enterprises that often utilize spend-management solutions from providers like American Express or other bespoke systems rather than Ramp. Furthermore, Ramp has opted to share only market share percentages, not actual dollar amounts spent, which would provide an even more granular view of revenue trends.
Despite these limitations, the data provides compelling indications of market sentiment and vendor preference. The "non-sticky" nature of enterprise AI spending is a significant takeaway. Businesses appear willing to frequently re-evaluate and switch between AI providers as new models are released, features evolve, and pricing structures are adjusted. This fluidity should prompt investors in both OpenAI and Anthropic to carefully consider the long-term sustainability and loyalty of enterprise customers. The competitive ebb and flow suggests that winning market share is not a permanent state but an ongoing battle fought on multiple fronts.
The Chronology of a High-Stakes Rivalry
The journey of OpenAI and Anthropic to their current market positions is deeply intertwined with the broader narrative of the generative AI boom.
-
OpenAI’s Initial Ascent and ChatGPT’s Impact (Late 2022 – Early 2023): OpenAI burst into the mainstream consciousness with the public launch of ChatGPT in November 2022. Its user-friendly interface and impressive capabilities quickly captivated both consumers and businesses, establishing OpenAI as the undisputed early leader. The subsequent release of GPT-4 further solidified its technological lead, making it the default choice for many developers and enterprises exploring AI integration. This period saw OpenAI enjoy a "runaway lead" in both mindshare and, presumably, market share.
-
Anthropic’s Emergence and Focus on Safety (2021 – Mid-2023): Anthropic was founded in 2021 by former OpenAI researchers who left over disagreements regarding the safety and ethical development of AI. Their flagship model, Claude, gained traction by emphasizing safety, interpretability, and responsible AI practices. This focus resonated particularly with enterprises concerned about data privacy, compliance, and mitigating AI-related risks. Anthropic positioned itself as a "safer" alternative, attracting businesses in highly regulated industries or those with stringent ethical guidelines. Over time, Claude’s performance improved significantly, making it a compelling option for a growing number of enterprise use cases.
-
The May Crossover (May 2024): This was a pivotal moment. Ramp’s data showed Anthropic surpassing OpenAI in market share among its business users, hitting 41% to OpenAI’s 39%. This indicated a successful penetration by Anthropic into the enterprise segment, leveraging its safety-first approach and increasingly capable models. It suggested that businesses were actively seeking alternatives and valuing the distinct offerings of different providers.
-
OpenAI’s Q3 Resurgence (July – Present): The latest data from Ramp points to a reversal of this trend. OpenAI is now showing faster growth in Q3. Ara Kharazian attributes this resurgence to the performance of what he refers to as "GPT-5.6 Sol," which he posted on X is "really good, increasingly the choice for developers." This suggests that OpenAI’s continuous innovation and model improvements are paying off, appealing to developers who prioritize raw performance and cutting-edge capabilities.
-
Anthropic’s Recent Challenges and Strategic Considerations: While Anthropic continues to hold a larger share as of July, its growth rate has reportedly slowed in Q3. Kharazian noted that "Fable 5," Anthropic’s higher-end model tier, "disappointed both in adoption and real-world application given price + data retention requirements imposed by regulators." Fable 5, designed for more targeted and complex use cases, is known to be expensive. Furthermore, Anthropic faced some backlash when it informed Fable users that it must retain their data for 30 days. This policy, likely driven by internal safety protocols or regulatory interpretations, sparked "outrage" among some users, highlighting the critical importance of data privacy and transparency in enterprise AI adoption. While Fable 5 is aimed at specific, high-value applications, any perceived misstep in data handling or pricing can impact broader enterprise trust and adoption across their product portfolio.
Factors Influencing Enterprise AI Adoption: Beyond Raw Performance
The shifting market shares underscore that enterprise AI adoption is influenced by a multifaceted set of factors, extending beyond just the raw capabilities of a large language model.
-
Performance and Innovation Velocity: OpenAI’s recent surge, attributed to "GPT-5.6 Sol," highlights the importance of continuous model improvement and the delivery of advanced capabilities. Developers and businesses are constantly seeking faster, more accurate, and more versatile models that can tackle increasingly complex tasks. The rapid pace of innovation means that a leader today might be challenged tomorrow if they do not maintain a high velocity of new releases and improvements.
-
Cost-Effectiveness and Pricing Models: The mention of Fable 5’s price points to cost as a significant barrier or accelerator. Enterprises, while eager to leverage AI, are also highly sensitive to return on investment (ROI). Pricing models (e.g., token-based, tiered subscriptions, enterprise licenses) and their perceived value for specific use cases play a crucial role. A model, no matter how powerful, might struggle with adoption if its cost structure is not aligned with the budget and value proposition for businesses.
-
Data Privacy, Security, and Compliance: The "outrage" over Anthropic’s 30-day data retention policy for Fable 5 serves as a stark reminder of the paramount importance of data governance in the enterprise space. Businesses handle sensitive customer, proprietary, and regulated data, and any AI vendor’s policies regarding data usage, retention, and security are scrutinized heavily. A strong stance on data privacy, clear transparency, and robust security protocols are non-negotiable for many enterprises, particularly those in finance, healthcare, legal, or government sectors. Providers that can offer enhanced privacy protections, such as zero-retention policies or on-premise deployment options, gain a competitive edge.
-
Developer Experience and Ecosystem Integration: OpenAI’s "GPT-5.6 Sol" being "increasingly the choice for developers" speaks to the importance of the developer experience. Ease of integration via APIs, comprehensive documentation, strong community support, and compatibility with existing tech stacks can significantly influence adoption. A robust ecosystem of tools, libraries, and third-party integrations makes it easier for businesses to deploy and scale AI solutions.
-
Ethical AI and Trust: Anthropic’s founding ethos around safety and ethical AI continues to be a differentiating factor. While the Fable 5 data retention issue presented a challenge, the underlying commitment to responsible AI remains attractive to many enterprises seeking to build trustworthy AI applications and avoid reputational risks.
Broader Market Context and Implications for the AI Industry
Ramp’s data also offers a crucial insight into the overall health and expansion of the enterprise AI market. Despite the intense competition for market share between OpenAI and Anthropic, the overall pie is growing. The percentage of Ramp’s business customers that pay for AI services has been steadily climbing, topping 50% in March and reaching nearly 56% by July. This indicates a robust and expanding market, where more and more businesses are integrating AI tools into their operations. This overarching growth suggests that even as these two giants battle for dominance, there is ample room for both to grow their revenue streams.
The dynamism observed in Ramp’s data is characteristic of a nascent but rapidly maturing technology sector. The competitive landscape extends beyond just OpenAI and Anthropic. Tech giants like Google (with Gemini and Vertex AI), Microsoft (with Azure AI and its OpenAI partnership), and Meta (with Llama) are also significant players, alongside a vibrant ecosystem of open-source models and specialized AI startups. This intense competition is a powerful driver of innovation, forcing all players to constantly refine their models, improve their offerings, and differentiate themselves on performance, price, safety, and ecosystem support.
For investors, particularly those eyeing the eventual IPOs of OpenAI and Anthropic, the data underscores both the immense potential and the inherent volatility of the AI market. The "stickiness" of enterprise spending, or lack thereof, suggests that valuations will need to account for rapid shifts in customer loyalty and the ongoing need for significant R&D investment to stay competitive. Long-term success will hinge not just on technological superiority but also on strategic pricing, robust data governance, and an ability to adapt quickly to evolving enterprise needs and regulatory landscapes.
Conclusion: A Race with No Permanent Leader
The latest intelligence from Ramp paints a vivid picture of a fiercely contested enterprise AI market where leadership is fluid and hard-won. OpenAI’s renewed vigor, fueled by model advancements, signifies its determination to reclaim its dominant position, while Anthropic’s strong existing share, despite recent challenges, demonstrates the enduring appeal of its safety-first approach.
The rapid shifts in market share, the increasing number of businesses adopting AI, and the continuous innovation from all players confirm that the enterprise AI race is far from over. There is no permanent leader in this domain; instead, companies must relentlessly innovate, address customer pain points around cost and data privacy, and build robust ecosystems to secure and maintain their competitive edge. As the market continues to expand and evolve at an unprecedented pace, the next quarter, or even the next month, could once again redefine who stands at the forefront of the AI revolution. The coming financial disclosures from OpenAI and Anthropic, once they near their IPOs, will undoubtedly provide further clarity on the revenue implications of these dynamic market share movements.







