Stability AI secures $232M Series B from music‑industry giants

Six Roles, One Stack

Stability AI has six open positions on Zero G Talent's live board, a move that intensifies competition for specialized AI engineers and prompts the firm to tighten its technical screening process. The roles span infrastructure, research, product, and people operations, signaling the company is staffing the full stack of an open-source generative AI lab. Zero G Talent's board data for Stability AI shows a salary band of $76,000–$118,000 with a median of $112,000 across salaried listings, though individual ranges vary by function.

The most technical opening sits in research: a Research Scientist, Professional Creative Workflows role listed as remote.

Two engineering roles target the audio stack. Product Design Engineer - Audio (United States) and Senior Product Engineer, Growth & Lifecycle Infrastructure - Music & Audio (Los Angeles or remote) sit on the product side.

Infrastructure gets a dedicated hire in IT Support Engineer (United States, $85,000–$120,000) and its junior counterpart, Junior IT Support Engineer (United States, $75,000–$103,000). At a lab running multi-thousand-GPU clusters, "IT support" means cluster health monitoring, driver firmware wars, and the on-call rotation that keeps training runs from silently diverging.

Rounding out the six is Technical HR Business Partner (United States). This is the only non-engineering role, but "technical" is the operative word.

Together the six roles form a coherent picture: Stability is hardening the operational layer beneath its open releases. The research scientist pushes the frontier. The audio engineers productize it. The IT engineers keep the compute alive. The HR partner ensures the next six hires clear the same bar. Every role requires evidence of open-source contribution — not as a nice-to-have, but as proof the candidate operates in the same default-public mode the company does.

How the Interview Loop Works

No public, step-by-step breakdown of Stability AI's interview loop exists. What exists instead is a broader industry vacuum: as of September 2026, "there is currently no widely accepted or standardised method for evaluating AI proficiency during recruitment," said Second Talent co-founder Elton Chan, and "traditional technical interviews are similarly ill-equipped to assess these capabilities."

That gap forces any company hiring at the model layer to invent its own signal. Stability AI's public commitments — open-weight releases, licensed training data, and a stated move toward "open training" where model development is shown in real time — suggest a screening bias toward engineers who can demonstrate reproducible experimentation, not just benchmark-chasing. A 2023 YouTube presentation from leadership emphasized that the team "squished" two billion images into generative models that powered four of the top ten avatar apps by December of that year.

Employers are moving toward behavioral probes, said People Matters in September 2026: "ask candidates to describe specific situations in which they have used AI, the problem they were trying to solve, how they applied the technology and what measurable outcome it produced." For Stability AI, that translates to the same: not GitHub stars, but merged PRs to diffusion transformer repos, audio codec work, or LLM fine-tuning pipelines that survived community review. The company's collaboration with 200 universities also implies a culture-fit screen for researchers comfortable publishing preprints, responding to public issues, and iterating in the open, a workflow distinct from the closed-lab cadence of frontier labs.

The Series B close announced August 25, 2026, brought total funding to $232 million with participation from Electronic Arts, Sony Music Group, Universal Music Group, and Warner Music Group. That investor roster signals a product roadmap where audio and video generation must clear IP clearance gates. Technical assessments likely include a practical exercise: given a licensed dataset slice, design a training run that respects attribution constraints and outputs a model card meeting transparency standards. Culture fit, meanwhile, hinges on whether a candidate treats "open" as a license condition or a development discipline. The company's own language, "no creative challenge too big, no timeline too tight," reads as a filter for engineers who have shipped under regulatory scrutiny, not just research deadlines.

Without a published interview rubric, applicants should prepare for a loop that weights reproducible artifact review over whiteboard coding, and that probes for multi-modal fluency across the 78-engineer team's surface area. The absence of a standard industry benchmark means Stability AI defines its own — and the candidates who pass will be the ones who can show, not tell, how they build in the open.

What the Pay Looks Like

Stability AI's two most recent postings, IT Support Engineer at $85,000–$120,000 and Junior IT Support Engineer at $75,000–$103,000, sit at the lower end of the company's own reported salary band. The four other open roles, Product Design Engineer - Audio, Technical HR Business Partner, Research Scientist for Professional Creative Workflows, and Senior Product Engineer for Growth & Lifecycle Infrastructure (Music & Audio), carry no public salary figures, leaving a gap for the core model-building positions the company says it is prioritizing.

Market data underscores the gap. As of August 2024, the average machine learning engineer commands $152,000 and the average data scientist $152,000, per Nexford's compilation of U.S. roles. Principal software engineers median at $164,000. Cloud architects and software architects both top $214,000. Generative AI specialization carries a documented premium: Indeed measured the average salary for any role requiring generative AI skills at $175,000 per year as of June 2024. Stability AI's posted IT support ranges fall 30–50 percent below those benchmarks, which is appropriate for infrastructure support but signals that the company's most competitive offers, equity-heavy packages for model researchers and audio product leads, remain undisclosed.

Geography complicates the picture. The Senior Product Engineer role lists Los Angeles or remote; the Research Scientist role is fully remote. New York metro data from Pace University shows mid-level AI roles at $128,000–$220,000 and senior roles at $224,000–$359,000 as of 2024. Remote hires at Stability AI may be calibrated to a national band rather than a coastal one, which would depress the top end unless the company applies location-agnostic pricing, a practice OpenAI and Anthropic have adopted for senior research roles.

Equity terms are absent from the public postings. Pre-Series B employees received options priced against a valuation that collapsed during 2023–2024 leadership turnover; the new round likely reset the 409A. Without a published strike price or refresh policy, candidates cannot model upside. Benefits details, such as health plan tiers, 401(k) match, compute stipends, publication budgets, and conference travel, are also missing from the board listings. Peer labs (OpenAI, Anthropic, Cohere) routinely offer $10,000–$25,000 annual compute credits and full conference coverage for research staff; Stability AI has not confirmed parity.

The table above blends the two verified Stability AI bands with the closest market analogs from Nexford (August 2024) and the Indeed generative AI aggregate (June 2024). It illustrates a clear pattern: the company's public numbers address support functions, while the roles that drive its open-source model roadmap, research scientists and senior product engineers in audio, carry no public compensation signal. Until Stability AI publishes bands for those positions or a candidate reaches offer stage, the only reliable anchor is the broader market, where generative AI expertise commands a $175,000 average and top-tier research hires routinely clear $250,000–$350,000 total compensation in coastal hubs.

A Lifeline in a Frozen Market

Stability AI's push arrives inside a labor market that has effectively bifurcated. Broad software hiring has collapsed — Indeed reports software development postings down more than half from their late‑2022 peak, and the information sector has shed 342,000 positions, an 11 percent decline, since that same peak. Yet the niche Stability is fishing in (engineers who have actually shipped open-weight models and contributed to public repositories) remains fiercely contested. The National Association of Colleges and Employers projects starting salaries for computer science graduates will rise nearly 7 percent year‑over‑year, a signal that demand for senior, specialized talent is decoupling from the entry‑level freeze.

That freeze is real and documented. Erik Brynjolfsson and Stanford's Digital Economy Lab found a 16 percent decline in early‑career employment across the most AI‑exposed occupations since ChatGPT's release — one in six jobs gone, with developers aged 22‑25 seeing a nearly 20 percent drop, one in five. BCG clocked software engineering headcount growth across the tech sector at barely 2 percent annually since late 2022, down sharply from the prior decade. McKinsey surveys show nearly a third of companies expect AI to shrink their employee base by at least 3 percent within a year. Hiring has slowed to 2010 levels — what economists now call a "big freeze", and worker confidence has cratered: just 28 percent of U.S. workers say it's a good time to find a job, down from 70 percent in 2022, with college graduates more pessimistic (19 percent) than non‑graduates (35 percent).

Against that backdrop, Stability's six roles read as a targeted play for the thin slice of the market that still moves. The company's first‑party board data confirms that band (median $112,000) for the two salaried roles posted in the past week, with an IT Support Engineer listed at that range and the junior role at $75,000–$103,000. Those figures sit above the median for generalist IT but below the premiums commanded by senior model‑builders at frontier labs, suggesting Stability is calibrating for contributors who value open‑source visibility and community standing over top‑of‑market cash.

Competitors feel the same pressure. Firms that use AI extensively grow employment 6 percent faster and sales 9.5 percent faster over five years, MIT Sloan found, but they also automate the tasks that once formed the entry ramp: call centers, agencies, offshore support. Salesforce cut roughly four thousand customer‑service roles after agents handled half of interactions; IBM eliminated two hundred HR positions via its "AskHR" agent. The first jobs to disappear are outsourced ones. That dynamic shrinks the pool of engineers who ever get to touch a model in production, making Stability's requirement for "deep model‑building experience and community contributions" a filter that excludes a growing share of the workforce.

Community perception cuts two ways. Stability's open‑source heritage — Stable Diffusion, Stable Audio, Stable Video, still carries weight in researcher circles, and a hiring wave framed around scaling that work signals commitment at a moment when several peers have tilted toward closed, API‑only releases. But the same community watches the broader contraction: more than a third of industry roles, roughly 2.2 million U.S. jobs, face agentic‑displacement risk per Morgan Stanley, and the "opportunities that never materialize" — Brynjolfsson's phrase for the quiet disappearance of first steps, are felt acutely in open‑source contributor pipelines. If Stability's screen selects only engineers who already have public model‑shipping credits, it reinforces a pipeline that is visibly narrowing.

The net effect is a talent market where Stability's six roles are both a lifeline for a specific profile and a mirror of the structural shift: fewer entry points, higher bars, and a premium on proven open‑source contribution that most early‑career engineers can no longer afford to build.

Three Hiring Waves

Stability AI's current push for six specialized roles marks the third distinct hiring phase in the company's short, volatile history. The first wave was not a hiring wave at all. In late 2022, Stability functioned primarily as a compute patron: it donated AWS cluster capacity to EleutherAI, the decentralized research collective that co-created the initial version of Stable Diffusion alongside Stability's own researchers. That relationship was collaborative, not contractual. EleutherAI's Stella Biderman later noted the arrangement preserved independence: "If we were fully funded by one tech company, that seems like a much bigger potential issue from our end." Stability's headcount then was lean, built around a core model team and a handful of infrastructure engineers keeping the training cluster alive.

The second phase arrived in early 2023, when Stability joined Hugging Face, Canva, former GitHub CEO Nat Friedman, and Lambda Labs to fund the EleutherAI Institute, a formal nonprofit spun out to sustain open-source language model work. Stability's contribution was again compute and capital, not headcount. The company's own roster grew modestly through 2023 as it productized Stable Diffusion for enterprise customers, such as HubSpot, Mercado Libre, and Stride Learning, but hiring remained reactive, tied to specific integration contracts rather than a roadmap.

Then came the crisis. By early 2025, TechCrunch reported that co-founder and CEO Emad Mostaque had "reportedly mismanaged Stability into financial ruin," triggering staff resignations, a collapsed Canva partnership, and investor alarm. The company that had once defined the open-source generative AI moment was bleeding credibility. The turnaround began in March 2025: new capital from Eric Schmidt and Napster founder Sean Parker, a new CEO in Prem Akkaraju, and the appointment of James Cameron to the board. That same month, Stability announced a thirty-fold speedup for Stable Audio Open on Armv9 chips, generating an 11-second clip in eight seconds, signaling a pivot toward optimized, device-local models.

The third phase, now underway, looks nothing like the first two. The August 2026 Series B, $76 million from Universal Music Group, Warner Music Group, Sony Music Group, and Electronic Arts, bringing total funding to $232 million, locked in entertainment-industry distribution before a single new model shipped. Akkaraju framed it bluntly: "As more and more professional creatives and businesses adopt generative AI to power their production pipeline, it's important that our models and workflows are available everywhere for builders to build and creators to create." The six open roles on Zero G Talent's board map directly to that strategy.

Contrast that with 2022, when the "team" was effectively a Discord server and a shared GPU quota. The current surge is deliberate, well-capitalized, and scoped to ship product — not just publish weights.

Working in frontier tech? Zero G Talent tracks the openings: see every open Stability AI role, browse frontier tech jobs, the companies hiring, and the people building the field.