Explore / Work with AI

How Founders Use AI · Everyday AI

Using AI Safely

The resources behind 9 questions people actually ask, then the questions themselves, each with a straight answer.

On the AI map, 29 October 2026 AI assistants AI browsers Avatars and voice

What founders use for using ai safely, ranked by how many answers lean on it.

  1. OpenAI's page on what is used for training by default and how Temporary Chat keeps a conversation out of history and training. help.openai.com Guide Data controls in ChatGPT OpenAI In 3 answers
  2. 1,700+ real court cases where fake AI citations got people sanctioned; the best cautionary tale on the internet. damiencharlotin.com Tool AI Hallucination Cases Database Damien Charlotin In 2 answers
  3. The straight answer from Anthropic, including how to toggle training off. privacy.claude.com Guide Is my data used for model training? Anthropic In 2 answers
  4. Survey finding that 24% of US AI users use an agent regularly and 32% have let AI act for them without final approval, which is... menlovc.com Article 2026: The State of Consumer AI Menlo Ventures In 2 answers
  5. Google's own notice on what it keeps and who reads it, including the line asking you not to enter confidential information a... support.google.com Guide Gemini Apps Privacy Hub Google In 2 answers
  6. The National Cyber Crime Reporting Portal, where you file a complaint about fraud online. cybercrime.gov.in Tool Cybercrime Portal Indian Cyber Crime Coordination Centre (I4C) In 2 answers
  7. Chakshu, the government page for reporting a suspected fraud call, SMS or WhatsApp message. sancharsaathi.gov.in Tool Sanchar Saathi Suspected Fraud Communication Reporting Department of Telecommunications In 2 answers
  8. Gives the 2025 cyber-fraud figures and the Rs 11,158 crore saved through quick reporting on 1930. dailypioneer.com Article When your voice is no longer proof The Pioneer In 2 answers
  9. A checklist of the visual and audio signs of a deepfake, written for Indian investors. tatamutualfund.com Guide Deepfake Investment Videos: How to Check Whether That Fund Manager Clip Is Real Tata Mutual Fund In 2 answers
  10. BSE's warning that a video of its chief giving stock tips was fake, and to trust only official channels. deccanherald.com Article BSE CEO Deepfake: BSE Warns Against Fake Stock Tips Deccan Herald In 2 answers
  11. The single most important security idea for anyone deploying agents. simonwillison.net Article The lethal trifecta for AI agents Simon Willison
  12. Anthropic's equivalent answer, including the commercial-products carve-out. privacy.claude.com Guide Is my data used for model training? (Anthropic Privacy Center) Anthropic
  13. Anthropic on letting Claude act in the browser without approving each step, with a safety classifier checking every action. claude.com Article Claude in Chrome is generally available Anthropic
  14. Explains privilege and work product in plain terms and cites United States v. Heppner, where a defendant's chats with public... harvey.ai Article The Rules of Privilege Still Apply When Lawyers use AI Harvey (its Chief Legal Officer)
  15. A worked demonstration of a hidden Reddit comment steering an AI browser into a signed-in email account. brave.com Article Agentic Browser Security: Indirect Prompt Injection in Perplexity Comet Brave (Artem Chaikin and Shivan Kaul Sahib)
  16. Anthropic's research note on how it trains and screens a browsing agent against hidden instructions. anthropic.com Article Mitigating the risk of prompt injections in browser use Anthropic
  17. Microsoft says not to give Copilot confidential or sensitive personal data, such as anything revealing health, religion or... support.microsoft.com Guide Privacy FAQ for Microsoft Copilot Microsoft
  18. WhatsApp's help page on what Meta receives when you chat with Meta AI and that it may use your prompts to improve its AI. faq.whatsapp.com Guide About Meta AI WhatsApp
  19. How Meta AI users posted intimate chats to a public feed without realising, and the warning Meta added afterwards. engadget.com Article Meta warns users to 'avoid sharing personal or sensitive information' in its AI app Engadget (Karissa Bell)
  20. Shared ChatGPT links turned up in Google search until OpenAI removed the feature: a reminder that Share means public. techcrunch.com Article Your public ChatGPT queries are getting indexed by Google and other search engines TechCrunch
  21. The security researchers who found DeepSeek's open database, with over a million lines of logs including users' chat history. wiz.io Article Wiz Research Uncovers Exposed DeepSeek Database Leaking Sensitive Information, Including Chat History Wiz Research (Gal Nagli)
  22. Where the Help improve our AI models switch sits in Claude on a computer and on a phone, and what turning it off does to past... privacy.claude.com Guide How do I change my model improvement privacy settings? Anthropic
  23. Deleted Claude chats leave its systems within 30 days; with training on, data can stay up to 5 years. privacy.claude.com Guide How long do you store my data? Anthropic
  24. The Training on conversation activity and Training on voice conversations switches, with the path on web, desktop and phone. support.microsoft.com Guide Microsoft Copilot privacy controls Microsoft

Show the next 24