Poetic, a startup focused on automating business workflows, stepped out of stealth mode on Wednesday with $50 million in funding from OpenAI. The investment signals a deepening push by the AI giant to build its own ecosystem of tools — a move that critics warn could stifle open-source AI development.
What Poetic plans to build
The company says it will use the capital to develop software that handles repetitive enterprise tasks — things like data entry, scheduling, and report generation. Rather than offering a generic chatbot, Poetic aims to embed AI agents directly into corporate systems, letting them act on structured data without constant human oversight.
Poetic hasn't released a product yet, and the team remains small. But the backing from OpenAI gives it immediate credibility in a crowded market where startups like Adept and Inflection have also raised billions. The difference: Poetic is tying its fortunes tightly to OpenAI’s models, which are not open source.
Why OpenAI is investing in a potential rival
OpenAI typically invests in companies that use its technology, and Poetic fits that pattern. But the $50 million check comes at a time when the company is under fire for restricting access to its most powerful models. By funding startups that rely on proprietary APIs, OpenAI strengthens its own platform — and makes it harder for open-source alternatives to gain traction.
Some in the AI community have already voiced concern. Open-source models like Meta’s Llama and Mistral’s offerings allow developers to modify and deploy them freely. OpenAI, by contrast, keeps its code closed and charges per token. Poetic’s reliance on that closed stack means enterprise customers who sign up are locked into OpenAI’s ecosystem from day one.
The broader shift toward walled gardens
The investment fits a wider trend. Over the past year, OpenAI has made a series of strategic moves — from licensing deals with publishers to custom chip development — that tighten its grip on the AI supply chain. Funding Poetic extends that control into the application layer, where most users actually interact with AI.
Proponents of open-source AI argue that this approach could slow innovation. Without access to underlying model weights, third-party researchers can’t audit safety or build on top of the technology. Enterprise customers, meanwhile, lose the flexibility to switch providers or fine-tune models on their own data.
Poetic hasn't commented on those concerns. The company is expected to release a beta version of its automation platform later this year.
