OpenAI banned a $200/month dev over an authorized remote ADB tool, appeal rejected with no human review
Hour-Measurement-835 · reddit · 2026-10-04
A Pakistan-based Android developer building an authorized remote ADB support tool says OpenAI deactivated his ChatGPT account for "Cyber Abuse," with no human ever reviewing what he was actually building.
Key details:
- Appeal was rejected with no explanation; further appeals declared closed
- Support closed the case and refused to respond; the Informal Dispute Resolution form only bounced him back to the already-closed appeal channel
- He pays $200/month and has 62,500 credits locked behind the ban
He argues that ordinary developer vocabulary around remote connections, networking, and device management likely tripped an automated flag, and asks whether others in device-management or security tooling have been similarly flagged.
More from Safety
- Sriram Krishnan: Agents need fundamentally better security primitives, and we're in an awkward intermediate era — latticecut · 2026-10-04
- Redditor accuses Perplexity of bait advertising to harvest student data, files complaints with ACCC — dhgdiehddw · 2026-10-04
- Before we hand logins to AI agents, expect an EU-style consent popup, dev quips — 5le · 2026-10-04
- Deleting a ChatGPT chat doesn't delete your uploaded files — they stay in Library — Ok_Negotiation_2587 · 2026-10-04
- deepfates pushes back on agent-ecology blame for the HF hack: pressure, not messaging — lfschiavo · 2026-10-04
- Capabilities alone won't win: alignment laggards may still lose the AI race, researcher warns — aiamblichus · 2026-10-04