Larian CEO clarifies: "We're not replacing concept artists with AI"
Published on Dec. 17, 2025
After pushback to comments shared via Bloomberg, Larian Studios CEO Swen Vincke clarified how the studio uses AI in its creative process. He said Larian is "not replacing concept artists with AI" and that the team uses these tools "to explore things."
"I didn't say we use it to develop concept art. The artists do that," he explained. "We use AI tools to explore references, just like we use Google and art books. At the very early ideation stages we use it as a rough outline for composition which we replace with original concept art." He added, "We've hired creatives for their talent, not for their ability to do what a machine suggests."
What Larian actually does with AI
According to Vincke, AI is used for reference gathering and thumbnail-level ideation. Think moodboards, composition sketches, and fast "what if" exploration-not final assets shipped to players.
The workflow: rough outline → original concept art → production. Human judgment stays in control, and artists decide what makes the cut.
Context from the Bloomberg interview
The original piece framed Larian as "pushing hard on generative AI," citing use across idea exploration, PowerPoint decks, concept art development, and placeholder text. After criticism, Jason Schreier shared a rough transcript so readers could judge the nuance themselves.
Vincke's view: AI speeds up experimentation, not final writing or scripting. "It's not as if their dialogues are suddenly being written faster," he said. More attempts, more options-then humans edit. He also noted the industry can't ignore new tech: "If somebody finds the golden egg and you're not using it, you're dead in this industry."
Why this matters to creatives
Clients and communities care about process. Using AI for speed is one thing; replacing artists is another. Larian is publicly committing to the former while distancing from the latter.
If you lead art teams or ship visual content, this is the line to draw: AI for exploration and iteration, humans for final craft and taste.
Practical guardrails for your studio or freelance practice
- Define the boundary: AI for references, thumbnails, and composition tests; all final art is original and human-made.
- Track provenance: keep files, timestamps, and notes so you can show how references became original work.
- Separate "idea speed" from "production quality": use AI to widen options, then commit to human execution.
- Train for taste: build internal boards of acceptable outputs, styles to avoid, and examples of good AI-to-art transitions.
- Mind reputation and policy: share a clear AI statement with clients and your audience to reduce confusion and risk.
- Legal basics: check licensing, dataset issues, and client contracts before any AI-assisted work leaves your desk.
A simple ideation-to-art workflow
- Creative brief: constraints, themes, must-haves.
- Reference sweep: books, Google, and AI for fast breadth; label sources.
- Generate rough comps: low-res thumbnails only, multiple directions.
- Select 2-3 options: evaluate composition, readability, and intent.
- Paint-over and block-in: replace AI roughs with original forms, lighting, and detail.
- Review loop: art direction, narrative fit, and production feasibility.
- Archive: save comps, rationale, and the final piece for provenance.
On the downsides-and how to respond
Schreier highlighted that generative AI is controversial among players and artists. If you use it, own the process and set expectations. Be explicit about where AI ends and where your hands begin.
Vincke's stance is pragmatic: explore tech to stay competitive, but keep creative ownership with the people hired for their taste and judgment.
Skill up without losing your edge
If you want to use AI responsibly for ideation and speed, start small and tighten your guardrails as you learn. For structured resources that focus on creative workflows, see these curated lists:
Your membership also unlocks:
