Safety and alignment

OpenAI restricts Voice Engine over synthetic-voice misuse risk

OpenAI disclosed that Voice Engine can clone a natural-sounding voice from a 15-second sample but withheld broad release, limiting testing to trusted partners with consent, disclosure, watermarking and monitoring requirements.

0 comments · 0 votesOpen discussion

Public discussion is readable by everyone. Sign in to comment, reply, or vote.

No comments yet. Start the discussion.

CURRENT ASSESSMENT · REVISION 1
AWAY FROM DOOM38confidence 95/100

Why it moved the index

The company demonstrated a potent impersonation capability but deliberately limited distribution and required several practical controls against misuse, preserving more human and institutional control than a public release would have.

AUDIT TRAIL

Assessment history

  1. R1
    Away 38 · confidence 95

    New March 2024 restricted synthetic-voice capability preview and safeguard decision.

    12 Aug 2026
SHARE THE FINDINGS

Share this page

DoomBench social sharing card for OpenAI restricts Voice Engine over synthetic-voice misuse risk.
  1. DoomBench assesses “OpenAI restricts Voice Engine over synthetic-voice misuse risk” as evidence moving away from doom, with magnitude 38 and confidence 95 out of 100 in the safety and alignment category.

  2. The DoomBench assessment of “OpenAI restricts Voice Engine over synthetic-voice misuse risk” is based on reporting from OpenAI and records the editorial rationale, source quality, attribution, and revision history.

  3. DoomBench summarizes “OpenAI restricts Voice Engine over synthetic-voice misuse risk” as follows: OpenAI disclosed that Voice Engine can clone a natural-sounding voice from a 15-second sample but withheld broad release, limiting testing...

    https://www.doombench.com/news/openai-restricts-voice-engine-over-synthetic-voice-misuse-risk-2024-03-29