Clash of Titans: How Enterprise AI Agents and SaaS Platforms are Set to Collide

When OpenAI launched Frontier in February, it was presented as a platform for enterprise AI agents. However, the true implication was a direct challenge to the revenue model that has supported the software industry for nearly two decades.

Frontier serves as a semantic layer that connects an organization’s existing systems, such as data warehouses, CRM platforms, and internal applications. This enables AI agents to operate with the same business understanding as a human employee. OpenAI refers to these agents as “AI coworkers” that can be onboarded, assigned identities, granted permissions, and evaluated for performance.

Early adopters of Frontier include Uber, State Farm, Intuit, and Thermo Fisher Scientific. The platform’s commercial goal is evident, with OpenAI aiming to increase enterprise customers’ share of revenue from 40% to around 50% by the end of the year.

How Frontier Enhances Enterprise Workflows

The rationale for Frontier addresses a common issue raised by CIOs in 2025 and beyond: deploying agents independently adds complexity rather than streamlining processes. Each new agent necessitates its own integration, data connections, and governance controls, leading to fragmentation at scale.

OpenAI’s solution is a shared business context. Instead of each agent developing its own understanding of the organization, Frontier offers a centralized layer that all agents can utilize. Fidji Simo, OpenAI’s CEO of Applications, highlighted this during the launch, drawing from her experience at Instacart.

Results from early implementations of Frontier are impressive. A global investment firm using Frontier agents in its sales process reduced salesperson time spent on administrative tasks by over 90%. A technology customer reported saving 1,500 hours per month in product development. At a major manufacturer, agents streamlined a production optimization process from six weeks to a single day.

Frontier fosters an open environment, accommodating agents built by OpenAI, in-house enterprise teams, and third-party providers like Google, Microsoft, and Anthropic. This openness not only serves as a design principle but also positions Frontier as a comprehensive solution rather than a vendor lock-in strategy.

Addressing the Seat-License Challenge

The underlying concern for established players is structural. The traditional per-seat license model, which has been immensely lucrative for SaaS companies, assumes software usage correlates with headcount. If an AI agent can replace a human employee accessing Salesforce, the justification for that seat license diminishes. Fortune directly acknowledged this fear, stating that platforms like Frontier could render SaaS software “invisible” and less valuable.

Notably, Salesforce’s stock has declined more than 27% this year, primarily attributed to fears of AI disruption rather than financial weakness. Despite solid Q4 FY2026 results, with revenue reaching $11.2 billion and annual recurring revenue hitting $800 million, the stock dipped after hours due to guidance falling below expectations.

Incumbents are adapting to this shift. Salesforce introduced the Agentic Enterprise License Agreement, a fixed-price, all-inclusive model for Agentforce, aiming to make consumption more predictable for enterprise buyers. ServiceNow has transitioned to consumption-based pricing for some AI agent offerings and partnered with OpenAI to embed frontier model capabilities into its platform. Microsoft also introduced consumption-based pricing alongside its per-user model for Copilot Studio.

This pricing pivot signifies a realization among these companies that the seat-license model is unsustainable in the face of agentic AI. The critical question is whether repricing alone is sufficient or if a fundamental architectural change is necessary.

Diverging Approaches to the Intelligence Layer

The strategic divergence in enterprise AI revolves around a key question: should AI agents be integrated within systems of record or positioned above them? Salesforce and ServiceNow advocate for the embedded model, arguing that agents are most effective when closely connected to data. They posit that CIOs will trust governance controls more readily from vendors already managing their workflows.

Marc Benioff, Salesforce’s CEO, has labeled Agentforce as the “operating system for the agentic enterprise.” ServiceNow positions its AI Control Tower as a centralized governance layer for all agents, regardless of origin.

In contrast, OpenAI, along with Anthropic’s Claude Cowork, supports the overlay model. Frontier operates above existing systems, utilizing open standards to link them rather than replacing them entirely. The premise is that enterprises should not need to overhaul their infrastructure to implement production-grade agents across their operations.

Both approaches offer advantages, presenting enterprises with genuine trade-offs. Embedded solutions promise tighter data control and quicker value realization within a familiar ecosystem. Overlay solutions offer flexibility and avoid the pitfall of agents restricted to a single vendor’s data.

Incumbents possess decades of institutional trust and established contracts, while OpenAI boasts superior model capabilities and a compelling argument for running the intelligence layer across the entire enterprise rather than a single product line.

Key Considerations for CIOs

Frontier is currently accessible to select customers, with wider availability anticipated in the coming months. Public pricing details are not available, with organizations encouraged to contact OpenAI’s enterprise sales team for information.

For CIOs, the decision-making process is not yet binary. Many large enterprises utilize Salesforce, ServiceNow, and Microsoft infrastructure concurrently. The immediate question revolves around whether Frontier serves as an orchestration layer connecting these systems or a competitive platform displacing them.

Denise Dresser, OpenAI’s chief revenue officer, succinctly encapsulated the current state of enterprise AI agents, highlighting the need for a straightforward way to leverage agents as internal teammates without necessitating extensive restructuring.

This gap is precisely what every player in this field aims to bridge. What sets Frontier apart is the company’s established enterprise relationships, deployment track record, and robust model capabilities. While SaaS incumbents hold an advantage in trust and data, the question remains whether this will be adequate in the evolving landscape of enterprise software throughout 2026.

(Photo by Austin Distel)

Explore More: Insights into OpenAI’s Enterprise Strategy: Unveiling the AI Sales Race

Curious about AI and big data insights from industry experts? Discover the AI & Big Data Expo happening in Amsterdam, California, and London. This comprehensive event is part of the TechEx series, co-located with other leading tech events such as the Cyber Security & Cloud Expo. Visit here for more details.

AI News is brought to you by TechForge Media. Browse upcoming enterprise tech events and webinars here.