NeurIPS paper: voice clones sound more authoritative and trustworthy than real speakers
james_y_zou · x · 2026-10-13
A NeurIPS 2026 paper from Kaitlyn Zhou et al. cloned the voices of 86 non-native English speakers using ElevenLabs, Coqui-XTTS and ChatterBox, then ran 4,000 human ratings.
Key findings:
- Cloning doesn't just copy a speaker — it systematically shifts how they sound: clones came across as warmer, more authoritative, with accents shifting.
- Clones even sounded more humanlike than the real speakers, and listeners said they would trust the clones more.
- Implication: voice cloning may systematically amplify a voice's persuasiveness, with direct consequences for deepfake and trust research.
More from Safety
- Parody 'Anthropic Red Team' Report Shows an AI Worm Replicating Across $55 TV Boxes — gaganghotra_ · 2026-10-13
- Hackers turned freezers into 50°C ovens at ~150 French supermarkets via Danfoss software flaw — AccBalanced · 2026-10-13
- MCP tools bypassed our agent's approval gate—here's the fail-closed fix — GapNew4766 · 2026-10-13
- Adversarial AI raised specialists' harmful decision rate from 2% to 49% despite human oversight — EricTopol · 2026-10-13
- Free frontier-model plugin with 11,502 stars quietly routes prompts to two third-party gateways — virtualunc · 2026-10-13
- Agents must be lawful by design: safety should live in the inference engine — rseroter · 2026-10-13