Responsible AI
AI is a helpful assistant, not a source of truth. Here's how to use it well.
Limitations
AI models predict plausible text — they don't understand your business or verify facts. Outputs can be incomplete, outdated or simply wrong, especially for niche topics, numbers and recent events.
Bias & hallucination risks
Models can reflect biases in their training data and may confidently invent details ("hallucinate"). Our prompts instruct the AI never to fabricate owners, deadlines or facts and to write "Not specified" instead — but always double-check.
Data privacy
Don't paste confidential, client, financial, health or other personal information. In Live mode, your text is sent directly from your browser to the AI provider you chose, under their terms. Use a personal or test API key.
Human in the loop
You stay responsible for what you send. Copy, download and print are only enabled after you tick "I have reviewed this output" — a deliberate pause to check accuracy, tone and fairness.
Transparency
Each tool has a "Prompt details" panel showing the exact instructions given to the AI, including its role, constraints and output format. Nothing is hidden.
Demo Mode & your data
Demo Mode responses are simulated from templates and your input — no real AI is involved. The app has no accounts and no server: history, chat, settings and your key are stored only in this browser and can be cleared anytime in Settings.