Runway builds a hiring layer for AI video

Runway has launched DIFFUSE, a hiring marketplace for creative professionals who work with generative video. Independent artists, boutique studios, and production companies can publish profiles and portfolios, while brands, agencies, and studios can search for people experienced with models such as Runway’s Gen-4.5 and Aleph.

The marketplace addresses a recurring question from Runway’s enterprise customers: who can integrate generative video into an existing production pipeline and deliver finished work? DIFFUSE is open now.

• Talent: Independent creators, creative studios, and production companies.
• Buyers: Brands, agencies, and studios seeking generative media expertise.
• Core features: Searchable profiles, portfolio reels, and direct inquiries.

A directory grows into a marketplace

In 2024, Runway launched the Talent Network, a curated directory for creatives and companies using AI in media production. Applicants joined through Google Forms, and a dedicated job board remained on the roadmap.

DIFFUSE turns that early directory into a searchable product. Members can build profiles, publish reels, and receive inquiries directly from prospective clients. Production companies listed in the network include Tool, Builders Club, Silverside, Obsidian, Zukunft, [AI]imagination, and EDGLRD, Harmony Korine’s digital IP studio.

Job postings show early demand

Runway says it analyzed more than 1,000 open creative positions across nearly 400 companies. Seventeen percent explicitly mentioned AI skills or generative tools, and Runway was the most frequently requested video-specific tool in the sample.

Runway has not published the sample period, collection method, job categories, or criteria used to classify AI-related skills. Those omissions limit comparisons with the wider creative labor market, but the findings indicate that some employers have begun naming generative media experience in formal job requirements.

Runway links that demand to an emerging group of directors, artists, editors, and visual storytellers who adopted generative tools early. The job-posting analysis measures demand for those skills; it does not establish the technology’s net effect on employment, pay, or existing production roles.

Models need people around them

Runway has expanded beyond individual video models with Media Router, a developer platform for companies building generative media into their products and workflows. Its customers include Adobe, Cloudflare, ElevenLabs, Expedia, Shutterstock, and Quora.

DIFFUSE extends that strategy into staffing. Companies adopting Runway can use the same vendor’s network to find people familiar with its tools, reducing training time and integration risk. A growing pool of experienced creators could also strengthen Runway’s position against Luma, Kling, OpenAI’s Sora, and other video platforms by increasing the cost of switching workflows.

Production teams gain a faster way to identify specialists who can carry work from model output through editing, compositing, sound, approvals, and delivery. Tool fluency alone does not establish production experience, so portfolio quality, credits, and references will remain central to hiring decisions.

What each side gets

• Creative professionals: A portfolio channel aimed at buyers already seeking generative media skills.
• Brands and studios: Searchable access to artists and production companies with relevant tools in their workflows.
• Technical teams: A potential source of collaborators who understand how generative video fits into established production systems.
• Runway: A talent ecosystem that can encourage adoption and continued use of its models.

Key marketplace rules remain open

Runway’s launch announcement leaves several operational details unspecified:

• How applicants are reviewed and whether every profile is accepted.
• How search rankings, recommendations, and any paid placement work.
• Whether Runway charges creators or hiring companies.
• Whether the platform supports job listings, contracts, payments, or dispute handling.
• How portfolio ownership, client confidentiality, consent, and training-data provenance are handled.
• Whether profiles can emphasize work created with competing models or model-agnostic workflows.

Based on the announced features, DIFFUSE currently functions as a searchable portfolio and lead-generation network. Its long-term value will depend on the quality of its profiles, sustained buyer activity, transparent marketplace policies, and support for the mixed-tool pipelines common in professional production.