Less than a week after Anthropic released 11 open-source plugins that enable Claude Cowork to execute a series of automated processes in areas ranging from customer support to IT operations, OpenAI responded Thursday with a similar platform it calls Frontier.  

It said that its offering “gives agents the same skills people need to succeed at work: shared context, onboarding, hands-on learning with feedback and clear permissions and boundaries. That’s how teams move beyond isolated use cases to AI coworkers that work across the business.”

Frontier works with existing systems, the announcement said, allowing customers to integrate their applications using open standards, which takes away the need to replatform. The new AI coworkers are accessible through “any interface, not trapped behind a single UI or application.”

It added that a number of existing customers, including Cisco, T-Mobile, and Argentinian financial institution BBVA are piloting Frontier, and HP, Intuit, State Farm, Thermo Fisher, and Uber are early adopters.

Frontier viewed as a logical next step

Jason Andersen, VP and principal analyst at Moor Insights & Strategy, said he is not surprised at the mainstream world’s excitement about Anthropic and OpenAI entering the space, noting, “they have already shown themselves to be disruptors and are now positioning themselves more directly in the SaaS and enterprise productivity space.”

The problem, he said, is that many of the platforms that the AI pure plays are trying to disrupt have embedded similar agentic technologies, so customers have already been exposed to toolsets like these in Microsoft Office, SAP, Slack, and other products that also offer integration and out of the box agents. He wondered what OpenAI and Anthropic can offer to displace incumbents that already have similar products.

He pointed out that the real question all of these incumbents, and the AI vendors, will need to ask themselves if they want to remain relevant in the future is how they will not only augment, but also leverage agents to transform the customer value proposition.

Thomas Randall, a research director at Info-Tech Research Group, said that OpenAI still remains the most popular model provider for enterprise AI deployments, and Frontier is the logical next step in ensuring its models can integrate across enterprise tools and management.

However, he pointed out, “this step is not market-leading, and OpenAI is starting to lose some of its first-mover advantage.” He noted that OpenAI’s competitors, such as Anthropic, have been much more proactive with agentic automation across business workflows; Anthropic’s Claude has especially gained traction among developers.

“Moreover,” he said, “large SaaS platforms that touch multiple departments in an organization, such as ServiceNow and Salesforce, are embedding their [own] AI agents across these integrated workflows, too — from supply chain to sales.”

He noted, “the question for enterprises will be: which provider will become your standardized orchestration platform for AI workflows? Will it be OpenAI? Or, more likely, will it be a platform such as ServiceNow, which may leverage OpenAI models but already forms the backbone of much of the enterprise technology stack?”

Arun Chandrasekaran, distinguished VP analyst at Gartner, observed that OpenAI’s Frontier AI platform signifies an increased focus on enterprise clients. “The vendor wants to expand its footprint beyond models and ChatGPT and become a platform for architecting, orchestrating and governing AI and agents,” he said, pointing out that the most immediate benefit for AI leaders is quicker time to value for organizations already invested in OpenAI’s products. However, “this is predicated on OpenAI delivering a cohesive AI platform with robust governance controls and deep integration into enterprise workflows.”

There are risks in making a large platform bet on OpenAI, he added, including outsized dependency on a single strategic supplier in a fast-changing AI landscape, and significant upfront investment with uncertain payoff.

A place for both platforms in the enterprise

However, Sanchit Vir Gogia, chief analyst at Greyhound Research, sees a place for both OpenAI Frontier and Claude Cowork in the enterprise: “They’re not variations of a single idea. These are two fundamentally different products solving completely different problems inside the enterprise.”

Frontier, he said, “is all about orchestration. Think of it as the control layer, the connective tissue that makes a fleet of AI agents usable, governable and, most importantly, dependable. It gives these agents structure. They’re not just tools; each agent has identity, purpose, permissions, and memory. Everything is logged, measured, and controlled. That’s how you go from pilots to production at scale.”

Claude Cowork is different, said Gogia. He described it as “a doer. It’s local, fast, and self-contained. It acts like a highly skilled junior team member that can take on end-to-end work when equipped with the right plugins. Those plugins give it role-specific intelligence. … But Cowork operates in a silo. Each instance runs on its own; there’s no shared state, no centralized policy, and no cross-agent awareness. That’s fine at small scale, but it gets messy fast when you try to run 20 or 50 of them across an organization”

Therefore, he said, the two platforms are not in conflict, they’re complementary. “Cowork handles task-level automation. Frontier handles coordination, governance, and scale. … Deploying them together is where the real power lies.”

He said he views Frontier as a signal of real change in enterprise AI: “[It] is OpenAI stepping squarely into the enterprise infrastructure world. It’s a platform to run and manage AI across your business the way you run and manage applications or services.”

That platform, said Gogia, “addresses a real bottleneck we’ve been watching for the past year. Enterprises aren’t struggling with AI models. They’re struggling with deploying agents reliably, safely, and consistently. Everyone’s got an AI pilot somewhere, but few can say those agents are integrated into the business. That’s the velocity gap — and Frontier is meant to close it.”