Addressing Antigravity Bans and Reinstating Access

Original Article ↗ Hacker News Discussion ↗

Perception of Google’s Response and Ban Reversal

  • Many see Google’s reinstatement of access as “the right way” to resolve the Antigravity/Gemini bans, especially compared to Anthropic’s handling.
  • Others argue this doesn’t fix the underlying pattern: opaque, automated bans with little or no appeal, and a long-standing lack of real customer support.
  • Some note the bans applied only to Antigravity/Gemini tools, not full Google accounts, and push back on narratives claiming total account lockouts.

Subscriptions, Token Use, and (Anti)competition

  • Strong disagreement over whether subscribers should be free to use their quota via any client (e.g. Antigravity, gemini-cli, OpenClaw).
  • One camp: locking discounted tokens to first‑party tools is anticompetitive and suppresses third‑party agents; if resale is the problem, ban resale explicitly.
  • Other camp: subscriptions are discounted precisely because they’re limited to Google’s UX, telemetry, and expected human-scale usage; heavy automated use should be on metered API pricing.
  • Analogy debates: some compare this to wanting Netflix content in VLC / right-to-repair; others say you knowingly traded flexibility for a cheaper, ToS‑encumbered bundle.

OAuth Piggybacking, Headless Modes, and Policy Ambiguity

  • Core violation: using third‑party tools/proxies to “harvest or piggyback” Antigravity/Gemini OAuth tokens (often via OpenClaw) to exhaust quotas.
  • Disagreement over whether using one’s “own” OAuth token in other software can fairly be called “stealing” or “shady.”
  • Multiple commenters highlight a gray area: tools explicitly support “headless”/automation, yet bans target automated third‑party harnesses; it’s unclear where acceptable headless use ends and ToS violation begins.
  • Some suspect intentionally vague rules to curb distillation/abuse and push users toward higher-margin API usage.

Account Consolidation and Digital Identity Risk

  • Many view tying experimental AI products to primary Google accounts as dangerously risky, given Gmail’s role as de facto digital identity.
  • Even if only AI features were actually cut off, people fear scenarios where an AI-related ToS issue cascades into email, photos, or SSO access.
  • Large subthread advocates owning a domain and using independent email providers to avoid “digital death sentences.”

Broader Concerns About AI, Vendors, and Power

  • Commenters extrapolate: if AI “employees” are centrally controlled by a few labs, those labs (or governments/funds behind them) effectively gain veto power over businesses.
  • Raises questions about reliability, due process, and whether private firms should wield this kind of leverage over users and even governments.