From the first launch of the first chatbot, named ELIZA, in 1966, yes, you read that right, 1966, at MIT by Joseph Weizenbaum, until today, when chatbots like Gemini, Claude and ChatGPT have become part of our daily lives, the progress has been revolutionary. The big bet of the AI revolution is that it will increase productivity while reducing costs. So far, this is still a narrative. AI needs to prove it can actually deliver the future that its funders have promised to the world.
The question is how people use chatbots, and how companies like OpenAI use their data to train the models. Can you trust them? This is actually a rhetorical question. The incident with the app "Chat & Ask AI" gave us the answer. Chat & Ask AI, with over 50 million users, had 300 million messages from 25 million users exposed due to a misconfigured cloud database that was left publicly accessible without any protection. The app connects to AI models from OpenAI, Anthropic and Google, and a security researcher found that 103 out of 200 apps scanned had the same vulnerability. People were treating these chatbots like private journals, and their most vulnerable conversations got exposed because someone failed to lock the door. You can read more about it here: Malwarebytes
Cirflect has a different philosophy. We don't use your data for any reason. Your conversations stay local, never shared with third parties. Your data belongs to you and stays yours. Cirflect's AI runs locally on our own server using open-source technology. Your journal entries are stored in a database that never leaves that same server. No cloud services, no third-party access, no exceptions.
Closing this article, no AI can replace real human connection. If you are going through something difficult, talk to your friends, your family, or a health professional. A chatbot has no feelings and no real understanding of your world. Cirflect's AI is here to help you reflect and write, not to replace the people who actually care about you.