MarketScreener: OpenAI and Anthropic pivot from model sales to packaged implementations

OpenAI and Anthropic have spent May 2026 building businesses around getting AI systems into production, not just selling access to models. OpenAI said on May 11 it launched the OpenAI Deployment Company, a unit built to help customers deploy AI in “real-world environments,” while Anthropic said on May 4 it formed a new AI services company with Blackstone, Hellman & Friedman and Goldman Sachs to bring Claude into core operations. That activity gives shape to the argument in a recent MarketScreener piece: the competition is moving from raw model sales toward packaged implementation. The companies’ own announcements show them adding customer engineering, governance tooling, partner channels and cloud distribution around the model layer. ### What did OpenAI actually launch? OpenAI said on May 11 that the OpenAI Deployment Company is designed to help organizations “build and deploy AI systems they can rely on every day across their most important work.” The company said the effort will extend its use of Forward Deployed Engineers, or FDEs, who work inside customer environments on workflow redesign, infrastructure changes and production deployment. (openai.com) Denise Dresser, OpenAI’s chief revenue officer, wrote on April 8 that customers were asking for company-wide AI agents, enterprise deployment and measurable business results, framing that demand as the “next phase of enterprise AI.” OpenAI has also tied that push to product distribution through ChatGPT Enterprise, Codex and Frontier, according to its enterprise materials. (openai.com) ### How far has Anthropic gone beyond the API? Anthropic said on May 4 that it was creating a new AI services company with Blackstone, Hellman & Friedman and Goldman Sachs to work with mid-sized companies across sectors. (openai.com) Anthropic said the organization would help customers bring Claude into “their most important operations,” rather than limiting the relationship to model access alone. Anthropic has also built a broader enterprise layer around Claude. (openai.com) Its Claude Enterprise product page says the offering includes governance, data controls and admin infrastructure for organization-wide deployment, and its Claude Partner Network announcement said Anthropic would commit an initial $100 million in 2026 to training, technical support and joint market development for partners helping enterprises adopt Claude. ### Where do Google and the cloud platforms fit in? (anthropic.com) Google Cloud has become one of the clearest distribution channels for Anthropic’s enterprise push. Google Cloud says customers can use Claude through its Model Garden and Gemini Enterprise Agent Platform to build “production-ready AI applications and agents” on managed infrastructure. Those listings matter because they show how implementation now includes hosting, governance, throughput options and integration with broader enterprise stacks. (anthropic.com) MarketScreener’s framing about packaged deployments is consistent with that setup: the model is being sold with the platform, operational controls and customer support needed to run it inside large organizations. ### Why are the companies adding services now? OpenAI and Anthropic both describe enterprise adoption in operational terms rather than benchmark terms. (cloud.google.com) OpenAI said successful deployment requires identifying high-impact use cases and redesigning workflows around them, while Anthropic’s materials emphasize governance, security controls and support for enterprise-wide rollout. PwC, in an expanded partnership announcement with Anthropic last week, said Claude was already running in production across insurance underwriting, HR transformation, cybersecurity and mainframe modernization. (cloud.google.com) Anthropic quoted PwC U.S. Senior Partner and CEO Paul Griggs as saying clients were moving from exploration to “enterprise-wide impact.” ### What should readers watch next? May 2026 gave investors and customers several concrete markers to track: OpenAI’s Deployment Company launch on May 11, Anthropic’s services company announcement on May 4, and Anthropic’s $100 million Claude Partner Network commitment for 2026. (openai.com) Those are the pieces to watch as both companies try to turn model demand into deployment revenue. Google Cloud’s Claude platform pages and OpenAI’s business announcements now provide the clearest public record of how that rollout is being packaged for enterprise buyers. (anthropic.com) Future updates are likely to appear first in those product and company channels. (cloud.google.com) (openai.com)