Questions people actually ask.

Ask Safely is an AI assistant that deletes by default and remembers by choice. Powered by Anthropic's Claude, it gives you top-tier AI with a key difference: your conversations auto-delete after 8 hours, your data is never used to train AI, and you're building a portable Memory Profile that personalizes your AI across every platform. Think of it as your AI passport. You own it, you control it, and you take it with you.

Ask Safely runs on Anthropic's Claude models: Haiku 4.5 and Sonnet 4.5 for everyday questions, and Claude Opus 4.6 on Expert. The app picks the right model for each question automatically, a faster one for simple things, a more powerful one for harder or higher-stakes questions. Expert subscribers can also choose a model manually. You don't need to know anything about AI models. Just ask.

Yes. When your question needs current information, Ask Safely automatically searches the web using Brave (a privacy-focused search engine), finds the most relevant results, and gives you a sourced answer. You'll see citations in the response so you can check the original sources yourself. You don't need to turn anything on. Search activates on its own when it's needed.

Ask Safely is available on iPhone and iPad (App Store), Android (Google Play), and web browser at app.asksafely.ai. Your account syncs across all your devices.

Ask Safely is built by SafeLife Inc., founded by Mike Ebener. Mike previously led product at Walmart, scaling their last-mile delivery business. The team includes engineers across iOS, Android, and web, a UX researcher, a fractional CTO, and a UX designer. SafeLife is a Delaware C-Corp. You can reach the team at mike@asksafely.ai.

When you chat with tools like ChatGPT, your conversations can be used to improve their AI by default. Your words, your business ideas, your personal details become part of the training data for future models. Ask Safely never does this. The AI models are already trained before they reach you. Nothing you type ever goes back into making the AI smarter or shaping what it tells someone else. Your conversations stay yours.

Every message is encrypted with AES-256, the same standard banks and government agencies use to protect sensitive data. Each message gets its own unique encryption key. Your email and stored data are encrypted the same way. Your password is protected with industry-standard one-way hashing, meaning even we can't recover or read it. On top of that, chats auto-delete after 8 hours. That's a permanent, hard delete from the database. Not hidden in an archive somewhere.

Yes. When the countdown timer hits zero, the conversation is permanently removed from the database. It's not moved to a trash folder, not archived, and not recoverable. The same goes for instant deletion. When you swipe left to delete a chat or click delete on web, it's gone immediately.

No. Ask Safely does not sell, share, or provide your data to any third party. Your conversations are not used for advertising, profiling, or training. We make money from subscriptions, not your data.

You can ask about anything without worrying about a permanent record. Because chats auto-delete and your data is never stored long-term or used for training, there's no lasting footprint from questions about health symptoms, legal situations, finances, relationships, or anything else. That said, Ask Safely is an AI assistant, not a doctor, lawyer, or financial advisor. Always treat answers as a helpful starting point.

No. The privacy protections are completely separate from the AI's ability to answer your questions. Ask Safely uses the same top-tier Anthropic models available anywhere. The auto-deletion and encryption happen after you get your answer. They don't affect the quality of the response at all. Full AI capability, full privacy. No trade-off.

Ask Safely is completing its SOC 2 Type 2 certification through Vanta, with certification expected by mid-2026. Our entire platform runs on AWS infrastructure, which maintains SOC 2, PCI DSS, and HIPAA certifications, giving us a strong security foundation while we complete our own independent audit.

By default, every chat auto-deletes 8 hours after your last message. You'll see a live countdown timer showing how much time is left. If you want a conversation to last longer, extend it: swipe right on mobile to keep it 14 days, or use the extend buttons on web for 24 hours or 14 days. After 14 days, it's permanently deleted. You can also delete any chat instantly at any time.

When you save (extend) a chat, it moves from the 8-hour default to a 14-day window. You'll see its own countdown timer in your chat history so you always know exactly when it will disappear. During those 14 days you can re-read it, continue the conversation, or delete it early. After 14 days, it's permanently gone.

On mobile apps, swipe left on a chat to delete it instantly. Swipe right to extend it to 14 days. As you start swiping, a label shows which action you're about to take so you don't accidentally do the wrong one. On the web app, the same actions are available through buttons.

No. Starting a new chat moves your current conversation to your chat history. It doesn't delete it. Your previous chat keeps its countdown timer and stays accessible until it expires or you delete it.

No, and that's by design. Deletion is permanent and immediate. There's no trash folder, no undo, no recovery. If you have a conversation you might want to come back to, extend it to 14 days before you're done with it.

Paid subscriptions. The Essential tier is free and always will be. The Expert tier is $20/month or $200/year and includes our most advanced AI, Deep Research, voice, document upload, and a Memory Profile you fully control. The first 100 Expert seats are Founding Memberships: $100/year, locked for life. We chose this model deliberately. When revenue comes from subscriptions instead of data, our incentives line up with yours.

The free Essential tier is genuinely free with no hidden data collection. You get unlimited AI chat, automatic web search, auto-deleting chats, the ability to save and extend conversations, guided prompts, follow-up suggestions, and cross-device sync. It handles about 80% of what most people use AI for. We offer it because trust is earned, and it's funded by the paid Expert tier.

Everything in Essential, plus Claude Opus 4.6, Anthropic's most powerful model, with an extended-thinking toggle and manual model selection. Expert also adds Deep Research, voice conversations, dictation, document upload and analysis, and private chat sharing links. A Memory Profile you own and control arrives in August. Founding Members ($100/year, locked for life, 100 seats) get every Expert feature first and a direct line to the team: shape features, test early builds, and help design your AI passport before it goes public.

Your chats auto-delete by default, so there's nothing to lose there. If you've built a Memory Profile on Expert, you'll be able to export it anytime. We've designed the product so you're never locked in.

We don't run ads and have no plans to. When AI recommends a product or steers a conversation toward a brand, how would you know? Unlike traditional ads, AI-generated recommendations can be invisible, woven into answers you're trusting as objective. We'd rather make money from subscriptions and keep your conversations completely clean.

Deep Research is Expert's tool for questions one search can't settle. Ask something big, like comparing Medicare plans or whether to refinance, and Ask Safely breaks it into steps, runs the searches, reads the sources, and writes a cited report you can download as a PDF. You watch its progress live. It runs on Claude Opus 4.6. Early access for Founding Members now; included with Expert.

No. Guest Mode lets you try Ask Safely right away. No signup, no email, nothing. You get up to 10 messages a day with the same AI, web search, auto-deleting chats, and guided prompts. If you like it and want unlimited messages, you can create a free Essential account. Any conversations you started as a guest carry over.

Whatever's on your mind. Ask Safely shows guided prompt suggestions when you open a new chat. Categories like "Help me write," "Help me learn," or "Search a topic," each expanding into specific ideas. After each response, you'll also see follow-up suggestions that help you dig deeper. These prompts are there every session, not just your first time.

Yes, with Expert. Upload Word docs, PDFs, spreadsheets, and images (.docx, .pdf, .csv, .xlsx, .png, .jpg, .gif) and the AI reads, analyzes, and discusses them with you. Useful for reviewing a contract, making sense of a spreadsheet, or getting feedback on your writing. Uploads are encrypted and deleted on the same timers as your chats. Early access for Founding Members now; included with Expert.

Yes, with Expert. Voice-to-voice lets you talk with Ask Safely and hear spoken replies. Dictate turns your speech into text in the chat box. Voice gets the same privacy protections as text: encrypted, never used for training, deleted on the same timers. Early access for Founding Members now; included with Expert.

Yes, with Expert. Generate a private link to any conversation and set when the link expires. The person you share with sees the chat without needing an account. Early access for Founding Members now; included with Expert.

Memory and agents. In August, the Memory Profile goes live: a transparent, portable record of what you choose to save, set up through a short voice interview. Alongside it come guided agents for the jobs you actually do. Researcher, Writer, Troubleshooter, and Lookout arrive first, with more to follow. Founding Members test everything early. Public launch is September.

When the AI uses web search, it includes citations directly in the response. You can see which sources informed which parts of the answer and click through to verify. Below the response, you'll also see which AI model was used and whether search was involved.

For most conversations, it doesn't need to, and that's the point. About 80% of what people ask AI are one-off questions. For the things that matter, you choose what to save to your Memory Profile: your dietary needs, your kids' names, your work context, your communication preferences. You see everything. You edit anything. You take it with you to any AI that supports the open MCP standard, including Claude, ChatGPT, and Gemini. It's your AI passport. Founding Members are helping design it right now.

Yes. On iOS, you can enable Face ID or Touch ID in settings. The app will require biometric authentication to open, adding an extra layer of privacy on top of your account login.

Be specific. Instead of "help me write an email," try "help me write a polite but firm email to my landlord about a broken heater." Mention who the audience is and what tone you want. If the first answer isn't quite right, tell the AI what to change rather than starting over.

For everyday use (writing help, searching, learning, and general questions) you won't notice a quality difference. Ask Safely uses Anthropic's latest Claude models, which consistently rank among the best available. What you will notice: Ask Safely doesn't collect your data, doesn't build a hidden profile, and gives you a memory layer you actually own. One that works across any AI platform, not just ours.

The difference is what happens to your data, and who controls your AI memory. ChatGPT stores your conversations by default and builds a hidden profile you can't fully see, edit, or take with you. Ask Safely auto-deletes chats after 8 hours (permanent deletion, not archiving), never uses your data for training, and when you use the Memory Profile, you own and control everything in it. Your memory is transparent, portable, and works across AI platforms. Not locked to one company.

You can. Claude is great. Ask Safely adds three things on top. The privacy layer: auto-deletion, encryption, and no data training. The simplicity layer: automatic model selection, auto-triggering web search, and guided prompts. And the memory layer: a portable Memory Profile that works as your AI passport across platforms. Claude gives you the raw AI. Ask Safely handles the complexity, protects your data, and builds a personalized AI experience you own.

Either way works. Many people switch entirely because Ask Safely covers everything they need. Others keep ChatGPT for some things and use Ask Safely specifically for private conversations: health questions, financial planning, personal topics, or anything they'd rather not have stored permanently.

Because our business depends on it. We make money from subscriptions. If we violated your privacy, we'd lose our entire reason to exist. Beyond that, auto-deletion runs at the database level via TTL indexes (not application code), encryption uses AWS KMS with hardware security modules, and we're pursuing SOC 2 certification for independent third-party verification.