Embark Explains Arc Raiders' AI Voices, And The Reasoning Is Thin

Embark Studios has released a video explaining why Arc Raiders shipped with a mix of human and AI-generated voices, and the studio's reasoning has not landed well.

Arc Raiders arrived in October 2025 with a blend of recorded performances and synthetic speech, a decision that drew plenty of criticism. In March 2026 the studio said it had begun swapping the AI lines for human recordings. Roughly eight months on, a new video hosted by Danny O'Dwyer features chief creative officer Stefan Strandberg and audio director Bence Pajor tackling the subject head-on.

The arguments on offer

The main defense centers on the game's contextual callouts — the short pings players fire off to share information in the field. According to Pajor, capturing every variation would have been too tall an order, and he suggested performers would have been driven mad by the sheer volume. Critics were quick to note that voice actors have handled enormous scripts for decades, and that line counts are a production problem rather than grounds for automation.

Two of the game's voice actors also appear in the video, in a segment that reads as awkward, discussing whether they could claim full ownership of a performance when parts of it were not theirs. The video further frames the studio's silence during the backlash as a decision to focus on re-recording instead of talking — though re-recording dialogue is routine work, not something that only happens after a controversy.

Pajor also notes, apparently amused, that some players complained about AI even on lines that turned out to be human, implying the difference is undetectable and therefore unimportant. Strandberg, meanwhile, insists Embark is not trying to replace performers and that their work should be celebrated and given more room.

Why it matters

For players, the practical question is trust: if a studio cannot reliably say which lines are synthetic, credit and pay disappear somewhere invisible. Voice work is a livelihood, and trading real performance for generated audio is a net loss for the craft.

What would convince you that a studio is serious about protecting voice actors?