ChatGPT for Teams

Original Article ↗ Hacker News Discussion ↗

OpenAI’s new “ChatGPT for teams” plan introduces a mid-tier, business-focused version of ChatGPT with higher usage limits, admin controls, and a key promise not to train on customers’ data, at $25–30 per user per month. Commenters largely see it as classic SaaS segmentation between individual, team, and enterprise tiers, but debate whether privacy should be a paid feature, how much companies will actually pay per seat, and what this means for startups building on top of OpenAI or competing open‑source models. Many also note growing privacy concerns and the availability of opt-out mechanisms for training on individual accounts, alongside broader questions about OpenAI’s long-term business model and competitive moat.

Product & Pricing

  • ChatGPT Team is positioned as a multi‑seat, business-tier version of ChatGPT with:
    • Higher message caps than Plus, 32k context window, and admin console.
    • Ability to create/share custom GPTs within a workspace.
    • “No training on your business data or conversations.”
  • Pricing: $25/user/month annually or $30/user/month monthly, above the $20 individual Plus price.
  • Some see the delta as a “confidentiality tax” (pay more to avoid training and get higher limits).

Usage Limits & Capabilities

  • Message cap figures (e.g., “100 messages / 3 hours”) are mentioned but not clearly documented; official help pages only say “higher limits,” so exact caps are unclear.
  • Confusion over model context: web ChatGPT Team advertises 32k context, while API offers 128k in GPT‑4 Turbo; some wonder about tradeoffs.

Data Privacy & Training

  • A major theme: what tiers are used for training.
    • Team, Enterprise, and API are repeatedly described as not training on user data.
    • Plus and free ChatGPT default to training, but there are opt‑outs:
      • Turning off “chat history & training” (loses history, breaks some features like voice/plugins).
      • A less-visible privacy request form (“Do not train on my content”) that preserves history.
  • Opinions split:
    • Some trust OpenAI will honor explicit “no training” commitments due to legal risk.
    • Others distrust the company given broader data/copyright controversies; see this as paywalled privacy.

Strategy, Moat & Impact

  • Many see this as classic SaaS segmentation and an enterprise move: more seats, admin controls, SSO (only on Enterprise), network effects via shared GPTs.
  • Some argue OpenAI’s moat is shifting from pure model quality to:
    • Distribution, integrations, and org‑level workflows.
    • Lock‑in via ease of use and collaboration, akin to Slack.
  • Concerns surface that this crowds out startups building similar team UIs atop the API.

Use Cases & Limitations

  • Suggested internal uses: GPTs over HR docs, PRDs, basic coding help, shared knowledge.
  • Others note this version does not yet connect seamlessly to full codebases or large internal repos; it’s still document‑attachment / retrieval‑based.
  • Some see the marketing examples (bugfinding, charts, etc.) as weak or unrealistic versus existing tools (linters, analytics).

Other Feedback

  • Desire for “ChatGPT for Family” multi-user plans with web browsing, without per-person Plus pricing.
  • Significant confusion over the title “ChatGPT for teams” being misread as an integration with Microsoft Teams.
  • Mixed sentiment overall: recognition that the offer is attractive for businesses, but heavy skepticism around pricing, privacy defaults, and long‑term platform power.