The last year has witnessed a dramatic shift in the enterprise artificial intelligence landscape, with OpenAI progressively encroaching on Microsoft's perceived stronghold. Drugmaker Amgen Inc.'s migration from Microsoft Copilot to OpenAI's ChatGPT for thousands of its employees serves as a vivid illustration of this trend, highlighting the rapidly evolving dynamics—and the rising rivalry—between the industry partners. This development is not just a tale of product preference but a telling reflection of broader currents buffeting the world of AI-powered workplace tools.
At the core of this rivalry lies a fundamental paradox. Microsoft, having invested nearly $14 billion into OpenAI and making its GPT models central to countless features in Windows and Office, now finds its own enterprise AI ambitions threatened by its close partner’s standalone offerings. Copilot, Microsoft’s flagship AI assistant, was rolled out to great fanfare as the generative AI tide swept through corporate America. Yet, OpenAI’s ChatGPT, originally a consumer-facing chatbot, has stormed the enterprise market, buoyed by user enthusiasm and a reputation for cutting-edge capabilities.
While Microsoft built Copilot directly into its productivity suite—Outlook, Teams, Excel, Word—OpenAI focused on perfecting ChatGPT’s user experience, rapidly iterating on features and responding directly to the needs of a broad customer base. Amgen's senior vice president, Sean Bruich, reflected this sentiment succinctly, noting that OpenAI had succeeded in making their product “fun to use,” a qualitative edge not easily quantified but with significant implications for grassroots adoption within organizations.
Microsoft’s historic strength—deep corporate relationships and seamless integration across its operating system and productivity apps—remains a significant weapon. Gartner analyst Jason Wong points out that buying Copilot is often the “path of least resistance,” with IT administrators opting for tools that slot neatly into existing procurement and security frameworks. Yet, this legacy advantage can sometimes impede agility. As Wong observes, many organizations keep both Copilot and ChatGPT in pilot programs, closely watching usage stats and employee feedback as they determine which tool delivers the best return on investment.
Financial services giant New York Life Insurance Co., for example, is rolling out both tools to all 12,000 employees simultaneously, seeking to test adoption and outcomes in a head-to-head trial. This comparative approach, echoed in other firms’ strategies, signals a pronounced shift in how enterprises evaluate technology: value and flexibility, not sheer incumbency, increasingly drive decisions.
Microsoft’s Jared Spataro, chief of workplace AI initiatives, defends the process, emphasizing that not every model tweak is beneficial and that rigorous vetting ensures a consistent business-grade experience. His point is well-taken; enterprise customers demand security and reliability above all. However, as companies like Amgen and Bain & Co. experiment with both platforms, employees readily notice when ChatGPT possesses a fresh ability that Copilot lacks, amplifying a perception that OpenAI is simply more “cutting-edge.”
Bain’s experience bears this out. While roughly 16,000 staff regularly use ChatGPT, only about 2,000 primarily leverage Copilot—and then mostly within Microsoft-specific workloads, such as Excel. Ramesh Razdan, Bain’s CTO, concedes that Copilot is “improving,” but lags behind ChatGPT in terms of user enthusiasm and effectiveness.
Yet the gap may not last. OpenAI has recently introduced usage-based pricing models, potentially bringing per-employee costs lower than Microsoft’s standard rate. Early reports also indicate that OpenAI is granting discounts to customers who bundle purchases across its AI portfolio, creating further incentives for organizations to standardize on its tools. For large enterprises, particularly those wrestling with budget constraints and variable-seasonal user numbers, this flexibility may prove decisive.
Adam Lieberman, chief AI officer at Finastra, notes that while Copilot “eventually garners appreciation for its functionality,” early-adopter bias toward ChatGPT remains strong. That said, he also points out that many users are simply less familiar with Copilot’s full capabilities, suggesting that further evangelism and user education could help Microsoft claw back ground.
Yet OpenAI is not standing still. Beyond ChatGPT’s web interface, the startup has expanded its suite of paid products, targeting businesses, educational institutions, and individual power users alike. Critically, it has shown a willingness to partner with Microsoft’s cloud rivals and is investing in its own software and workflow integrations. The recent acquisition of Windsurf, an AI coding assistant that competes with Microsoft-owned GitHub Copilot, is both a practical move and a symbolic shot across Microsoft's bow.
For customers, this competitive dynamic offers benefit and risk. On the one hand, it promises a steady stream of innovation and, ideally, lower prices. The risk lies in fragmentation and “decision fatigue,” as IT teams juggle overlapping functionality and the need to standardize on platforms for compliance, audit, and procurement reasons.
For OpenAI, forging alliances outside the Microsoft empire—whether in cloud infrastructure, AI service delivery, or targeted software verticals—further reduces dependency and spreads risk. The more that large enterprises elect to contract directly with OpenAI for AI capabilities, the greater the potential for internal rivalry to spill into the open, complicating joint go-to-market efforts and potentially diluting focus.
What is clear is that neither party anticipates a quick resolution. Microsoft CEO Satya Nadella has made no secret of his ambitions, setting an explicit organizational goal to achieve “hundreds of millions” of Copilot users. Recent presentations to employees have emphasized big customer wins: Barclays, Accenture, and Volkswagen are among those reporting over 100,000 paid Copilot licenses each. Still, these victories may prove pyrrhic if the fastest-growing segment—knowledge workers keen to use cutting-edge AI—defaults to ChatGPT given the choice.
OpenAI, conversely, is racing to build comparable credibility. Its enterprise offerings tout SOC 2 compliance and robust privacy controls, but for many risk-averse customers, the lack of a proven track record at the scale and heterogeneity of modern enterprise deployments is reason to proceed cautiously. The balance of risk and reward will differ from company to company, but the ability to demonstrate clear superiority in both features and compliance could ultimately determine market share.
ChatGPT, meanwhile, thrives on the perception of novelty and frontier capability. Employees gravitate toward tools that help them work smarter, faster, or more creatively. As an example, Amgen’s experiment revealed meaningful gains in employee productivity and satisfaction after opening up wider access to ChatGPT, even as Copilot continued to serve well for routine tasks within the Microsoft ecosystem.
Gartner’s Jason Wong notes that the current moment is “kind of a showdown,” an inflection point likely to reverberate across software procurement, digital transformation, and IT governance for years to come. The ultimate winners—among both vendors and customers—will be those who navigate the balance between agility and stability, cost and capability, integration and flexibility.
The field remains unsettled. Many organizations, watching usage stats and gathering feedback, are taking a pragmatic approach, rolling out both Copilot and ChatGPT and letting the best tool win. Pricing, performance, integration, user satisfaction, and compliance will all factor into the next round of decisions.
What is certain is that neither “partner” intends to cede ground. The contest will likely spark further advances in AI’s utility, value, and usability at work. For CIOs, IT leaders, and frontline employees alike, the ability to test, evaluate, and adopt new solutions—without sacrificing security or strategic coherence—will be the defining challenge of the AI-powered enterprise era.
As this battle continues, observers should expect rapid change—and, quite possibly, more surprising alliances and splits—in the year to come.
Source: The Edge Malaysia OpenAI is nabbing Microsoft customers, fuelling partners’ rivalry
At the core of this rivalry lies a fundamental paradox. Microsoft, having invested nearly $14 billion into OpenAI and making its GPT models central to countless features in Windows and Office, now finds its own enterprise AI ambitions threatened by its close partner’s standalone offerings. Copilot, Microsoft’s flagship AI assistant, was rolled out to great fanfare as the generative AI tide swept through corporate America. Yet, OpenAI’s ChatGPT, originally a consumer-facing chatbot, has stormed the enterprise market, buoyed by user enthusiasm and a reputation for cutting-edge capabilities.While Microsoft built Copilot directly into its productivity suite—Outlook, Teams, Excel, Word—OpenAI focused on perfecting ChatGPT’s user experience, rapidly iterating on features and responding directly to the needs of a broad customer base. Amgen's senior vice president, Sean Bruich, reflected this sentiment succinctly, noting that OpenAI had succeeded in making their product “fun to use,” a qualitative edge not easily quantified but with significant implications for grassroots adoption within organizations.
Customer Dynamics: Familiarity, Flexibility, and Friction
Part of ChatGPT’s appeal is rooted in its first-mover advantage. Before Copilot’s formal enterprise debut, countless employees had already experimented with ChatGPT at home, growing comfortable with its conversational style and utility. When office deployments began, this existing familiarity gave OpenAI a crucial head start, allowing lines of business to advocate for its adoption from the bottom up. This grassroots pull contrasted sharply with Microsoft’s traditional top-down, IT-driven sales model.Microsoft’s historic strength—deep corporate relationships and seamless integration across its operating system and productivity apps—remains a significant weapon. Gartner analyst Jason Wong points out that buying Copilot is often the “path of least resistance,” with IT administrators opting for tools that slot neatly into existing procurement and security frameworks. Yet, this legacy advantage can sometimes impede agility. As Wong observes, many organizations keep both Copilot and ChatGPT in pilot programs, closely watching usage stats and employee feedback as they determine which tool delivers the best return on investment.
Financial services giant New York Life Insurance Co., for example, is rolling out both tools to all 12,000 employees simultaneously, seeking to test adoption and outcomes in a head-to-head trial. This comparative approach, echoed in other firms’ strategies, signals a pronounced shift in how enterprises evaluate technology: value and flexibility, not sheer incumbency, increasingly drive decisions.
Under the Hood: Tech Differentiation and Speed of Innovation
One of Microsoft’s biggest hurdles is perception—and, increasingly, reality—around its speed of innovation. Since both Copilot and ChatGPT draw from the same core OpenAI models, differentiation often comes down to two factors: integration and responsiveness to new model releases. Here, bureaucracy has at times hampered Microsoft. According to several insiders, it can take weeks for new OpenAI features to filter through to Microsoft products, as each update undergoes additional internal testing and vetting for stability and security.Microsoft’s Jared Spataro, chief of workplace AI initiatives, defends the process, emphasizing that not every model tweak is beneficial and that rigorous vetting ensures a consistent business-grade experience. His point is well-taken; enterprise customers demand security and reliability above all. However, as companies like Amgen and Bain & Co. experiment with both platforms, employees readily notice when ChatGPT possesses a fresh ability that Copilot lacks, amplifying a perception that OpenAI is simply more “cutting-edge.”
Bain’s experience bears this out. While roughly 16,000 staff regularly use ChatGPT, only about 2,000 primarily leverage Copilot—and then mostly within Microsoft-specific workloads, such as Excel. Ramesh Razdan, Bain’s CTO, concedes that Copilot is “improving,” but lags behind ChatGPT in terms of user enthusiasm and effectiveness.
The Pricing Wars: Value, Transparency, and Tactics
Pricing remains a fluid and contentious battleground. Microsoft has marketed Copilot at a headline price of $30 per user per month, significantly undercutting ChatGPT Enterprise, which Gartner analysts say has been offered at up to $60 per user per month. This price gap, coupled with Copilot’s integration advantages, has been a powerful lever in conversations with cost-conscious enterprises.Yet the gap may not last. OpenAI has recently introduced usage-based pricing models, potentially bringing per-employee costs lower than Microsoft’s standard rate. Early reports also indicate that OpenAI is granting discounts to customers who bundle purchases across its AI portfolio, creating further incentives for organizations to standardize on its tools. For large enterprises, particularly those wrestling with budget constraints and variable-seasonal user numbers, this flexibility may prove decisive.
Adam Lieberman, chief AI officer at Finastra, notes that while Copilot “eventually garners appreciation for its functionality,” early-adopter bias toward ChatGPT remains strong. That said, he also points out that many users are simply less familiar with Copilot’s full capabilities, suggesting that further evangelism and user education could help Microsoft claw back ground.
Ecosystem Effects: App Bundles, Integrations, and Competitive Threats
Microsoft’s history is replete with examples of upstart challengers—Zoom, Slack, Box—struggling to break into corporate environments already dominated by its integrated stack. The company is betting on a repeat performance here, counting on the power of unified tools and centralized administration. Copilot’s seamless embedding within Office 365 gives it—and by extension, Microsoft—an unparalleled reach into the day-to-day workflows of the global workforce.Yet OpenAI is not standing still. Beyond ChatGPT’s web interface, the startup has expanded its suite of paid products, targeting businesses, educational institutions, and individual power users alike. Critically, it has shown a willingness to partner with Microsoft’s cloud rivals and is investing in its own software and workflow integrations. The recent acquisition of Windsurf, an AI coding assistant that competes with Microsoft-owned GitHub Copilot, is both a practical move and a symbolic shot across Microsoft's bow.
For customers, this competitive dynamic offers benefit and risk. On the one hand, it promises a steady stream of innovation and, ideally, lower prices. The risk lies in fragmentation and “decision fatigue,” as IT teams juggle overlapping functionality and the need to standardize on platforms for compliance, audit, and procurement reasons.
Internal Tensions and Strategic Calculations
The OpenAI-Microsoft tightrope act has grown more precarious as competitive pressure mounts. Despite their deep financial and technical ties, the partners are increasingly at odds over control of AI’s future. Microsoft has hedged its bets, backing other AI startups and investing in building its own generative AI models, while expressing public reservations about OpenAI’s restructuring plans.For OpenAI, forging alliances outside the Microsoft empire—whether in cloud infrastructure, AI service delivery, or targeted software verticals—further reduces dependency and spreads risk. The more that large enterprises elect to contract directly with OpenAI for AI capabilities, the greater the potential for internal rivalry to spill into the open, complicating joint go-to-market efforts and potentially diluting focus.
What is clear is that neither party anticipates a quick resolution. Microsoft CEO Satya Nadella has made no secret of his ambitions, setting an explicit organizational goal to achieve “hundreds of millions” of Copilot users. Recent presentations to employees have emphasized big customer wins: Barclays, Accenture, and Volkswagen are among those reporting over 100,000 paid Copilot licenses each. Still, these victories may prove pyrrhic if the fastest-growing segment—knowledge workers keen to use cutting-edge AI—defaults to ChatGPT given the choice.
Security and Compliance: The Ultimate Enterprise Litmus Test
Security, reliability, and compliance are non-negotiable in enterprise software. Here, Microsoft’s decades-long investment in compliance certifications, data privacy, and global regulatory coverage acts as a moat. Particularly for highly regulated industries (financial services, healthcare, government), Copilot’s guaranteed alignment with existing IT and security practices will continue to win it favor—at least in the short to medium term.OpenAI, conversely, is racing to build comparable credibility. Its enterprise offerings tout SOC 2 compliance and robust privacy controls, but for many risk-averse customers, the lack of a proven track record at the scale and heterogeneity of modern enterprise deployments is reason to proceed cautiously. The balance of risk and reward will differ from company to company, but the ability to demonstrate clear superiority in both features and compliance could ultimately determine market share.
The End-User Experience: Usability Is King
In the race to capture enterprise mindshare, the crucial variable may be the everyday experience of the end user. As emerging evidence from Amgen, Bain, and others attests, tools that are perceived as intuitive, responsive, and genuinely helpful at streamlining “onerous tasks”—writing, research, data analysis—will win grassroots support regardless of top-down mandates. Copilot’s deep linkages with Microsoft software—Excel formula suggestions, Teams meeting summaries, document drafting—offer a consistent, if sometimes conservative, AI experience.ChatGPT, meanwhile, thrives on the perception of novelty and frontier capability. Employees gravitate toward tools that help them work smarter, faster, or more creatively. As an example, Amgen’s experiment revealed meaningful gains in employee productivity and satisfaction after opening up wider access to ChatGPT, even as Copilot continued to serve well for routine tasks within the Microsoft ecosystem.
Risks and Uncertainties: Lock-In, Fragmentation, and Future Directions
While the current rivalry stimulates competition and customer choice, several significant risks warrant close attention:- Vendor Lock-In: As enterprises bake AI deeper into core processes, the risk of technological lock-in rises. Committing heavily to either OpenAI or Microsoft today could constrain flexibility in the face of future market or technological shifts.
- Data Fragmentation: Using disparate AI tools across teams can hinder knowledge sharing, version control, and data governance, eroding the benefits of digital transformation.
- Security Gaps: Rapid adoption, particularly of newer or more “fun” tools, can sometimes outpace secure deployment, raising concerns about data leakage, model manipulation, or compliance failures.
- Feature Convergence: With overlapping capabilities, companies could find themselves overpaying for redundant features, driving up overall software costs without commensurate value.
What’s at Stake: The Broader Impact on AI Industry Trajectory
This unfolding drama has implications well beyond the balance sheets of Microsoft and OpenAI. It speaks to the broader tension between integrated platforms and best-of-breed tools—between the value of unified ecosystems and the promise of rapid, user-driven innovation. It also shapes the pace and trajectory of AI adoption in the enterprise, setting benchmarks for pricing, feature development, and support expectations.Gartner’s Jason Wong notes that the current moment is “kind of a showdown,” an inflection point likely to reverberate across software procurement, digital transformation, and IT governance for years to come. The ultimate winners—among both vendors and customers—will be those who navigate the balance between agility and stability, cost and capability, integration and flexibility.
Conclusion: The Future Is Fluid
For now, Microsoft retains formidable advantages: a vast installed base, robust compliance infrastructure, and unrivalled integration across workplace tools. OpenAI, unburdened by legacy systems and buoyed by user excitement, is matching these strengths with relentless innovation and a direct line to end users.The field remains unsettled. Many organizations, watching usage stats and gathering feedback, are taking a pragmatic approach, rolling out both Copilot and ChatGPT and letting the best tool win. Pricing, performance, integration, user satisfaction, and compliance will all factor into the next round of decisions.
What is certain is that neither “partner” intends to cede ground. The contest will likely spark further advances in AI’s utility, value, and usability at work. For CIOs, IT leaders, and frontline employees alike, the ability to test, evaluate, and adopt new solutions—without sacrificing security or strategic coherence—will be the defining challenge of the AI-powered enterprise era.
As this battle continues, observers should expect rapid change—and, quite possibly, more surprising alliances and splits—in the year to come.
Source: The Edge Malaysia OpenAI is nabbing Microsoft customers, fuelling partners’ rivalry