Wirestock raises 23 million dollars for creative AI datasets

Wirestock has raised $23 million in Series A funding to expand a business that supplies multi-modal creative datasets to AI labs, TechCrunch reports. The company, which previously helped photographers distribute work to stock platforms such as Shutterstock, pivoted in 2023 toward providing images, video, design assets, gaming content and 3D material for training and improving AI models. Wirestock says more than 700,000 artists and designers are signed up to its platform.

The round was led by Nava Ventures, with participation from SBVP, Formula VC and I2BF Ventures. Wirestock’s CEO Mikayel Khachatryan told TechCrunch the company initially sold its existing library “off the shelf” and then moved toward custom data requests for labs. It reports an annual run-rate revenue of $40 million and says it has paid out $15 million to contributors so far.

What is being financed is not only a library, but a production line. Wirestock requires applicants to complete an unpaid task as a quality check before being accepted, and uses a mix of AI and human review to evaluate work. It has retrained parts of its team to do detailed annotation and labeling for enterprise customers, and is building sales capacity to pitch to hyperscalers. The product is increasingly the workflow: recruiting, filtering, directing and verifying a large contributor base to generate the specific kinds of data a model maker wants this quarter.

The economics are straightforward. Model developers compete on outputs—how well a system draws, edits, or generates video—but the input bottleneck is curated training material that is legally usable and technically consistent. Data suppliers that can deliver targeted datasets on demand become a lever in that race, while contributors are paid per task rather than sharing in downstream model value. Wirestock says it allowed artists to opt out when it shifted its business, and that most switched over to providing data.

Wirestock employs 60 people and says the new funding will be used to hire research, engineering and product staff, and to build enterprise software for AI labs to collaborate on datasets. The company’s pitch rests on a simple claim: as models become more capable, the market for training material starts to look less like stock photography and more like industrial procurement.