You are currently viewing Privacy Check: ChatGPT, Gemini, Claude, and Perplexity for Personal Info
Featured image for Privacy Check: ChatGPT, Gemini, Claude, and Perplexity for Personal Info

Privacy Check: ChatGPT, Gemini, Claude, and Perplexity for Personal Info

  • Post author:
  • Post category:News
  • Post comments:0 Comments

Privacy Check: ChatGPT, Gemini, Claude, and Perplexity for Personal Info

Image sourced from tomsguide.com
Image sourced from tomsguide.com

If you’re typing sensitive stuff like health worries, money troubles, or family advice into an AI chatbot, what happens to those chats matters. Amanda Caswell at Tom’s Guide read the privacy policies for ChatGPT, Gemini, Claude, and Perplexity to see how they handle your data. She ranked them based on training use, retention, and human access. Spoiler: differences are huge.

ChatGPT: Data-Hungry by Default

OpenAI’s ChatGPT saves your chat history forever and feeds conversations into model training unless you opt out. You have to manually delete history or switch to Temporary Chat each time for privacy. Controls exist, but they’re not automatic. Caswell puts it last for personal info.

  • Training on chats unless opted out
  • Indefinite storage
  • No default private mode

Gemini: Solid Settings, But Humans Peek

Google’s Gemini ties into your account with good controls like disabling activity logging or 3-month auto-delete. Problem is, some chats go to human reviewers and stick around up to 3 years, even after you delete them. Fine for casual talks if you tweak settings first, ranks third.

  • Can turn off logging and set short deletes
  • Human reviews on select chats
  • Defaults to keeping data

Perplexity: Easy Controls Come Through

Perplexity starts with training enabled but has a clear opt-out toggle in settings, plus Incognito Mode for zero storage and memory clear options. Transparent and user-friendly, it lands second.

  • Simple training opt-out
  • Incognito for full privacy
  • Disable or wipe memory anytime

Claude: Hands-Off on Your Data

Anthropic’s Claude skips training on your chats unless you opt in. Deleted convos vanish in about 30 days, no routine human reviews unless you send feedback. Caswell calls it the top pick—no tweaks needed.

  • No training by default
  • Quick deletion
  • No automatic human checks

Quick Ranking for Trust

For personal info, Tom’s Guide order:

  1. Claude (max privacy, no effort)
  2. Perplexity (flexible tools)
  3. Gemini (works with setup, review risk)
  4. ChatGPT (opt-out required)

Bottom line: Claude wins if privacy’s non-negotiable. Perplexity’s great for control freaks. Always check settings, especially on ChatGPT or Gemini. For task picks beyond privacy—like Claude for coding—ZDNET weighs in on models like Gemini for explainers.

More stories at letsjustdoai.com

Seb

I love AI and automations, I enjoy seeing how it can make my life easier. I have a background in computational sciences and worked in academia, industry and as consultant. This is my journey about how I learn and use AI.

Leave a Reply