Users flock to open source Moltbot for always-on AI, despite major risks (arstechnica.com)

An open source AI assistant named Moltbot, formerly known as Clawdbot, has rapidly gained popularity, amassing over 69,000 stars on GitHub within just a month of its release. Developed by Austrian programmer Peter Steinberger, Moltbot enables users to interact with their personal AI assistant via popular messaging platforms like WhatsApp, Telegram, and Slack. Its proactive features, such as sending reminders and alerts based on calendar events, have drawn comparisons to the fictional AI assistant Jarvis from the Iron Man films. However, despite its promising capabilities, the software presents significant security risks, especially for casual users. The importance of Moltbot lies in its ambitious vision of an always-on AI that integrates seamlessly into users' digital lives. While it offers advanced functionality by leveraging commercial large language models like Claude Opus 4.5 from Anthropic, users must configure a server and manage authentication, which adds complexity and potential vulnerabilities. Additionally, running Moltbot can lead to substantial API costs due to its demand for frequent calls to AI models. As the AI/ML community watches its development, Moltbot highlights both the potential and pitfalls of open source AI tools in enhancing productivity while raising concerns about security and accessibility.