Follow ZDNET: Add us as your preferred source Google.
Important points of ZDNET
- Moltbot is gaining a lot of attention in the AI field.
- The tool’s developer describes it as “AI that actually does things.”
- Moltbot is best run in a silo like the 2024 M4 Mac Mini.
One of the biggest challenges facing AI developers today is building agents that are specific, practical, and have widespread use. Many agents may work well in narrow areas, such as managing email or debugging code. However, reliable AI systems that can autonomously handle a wide range of tasks remain a distant dream. On the other hand, deep-seated issues around illusions and security limit the adoption of agents in enterprises.
This is what makes the sudden viral popularity of Moltbot (touted by its developers as “an AI that actually does things”) so significant.
Also, is ChatGPT Plus still worth $20? I compared it to the Free, Go, and Pro plans – here’s my advice
Moltbot is being touted as an AI assistant that can manage almost every aspect of your digital life, including sending emails, managing your Google Calendar, and opening airline apps to check in for your next flight. However, like any AI assistant that requires access to your personal account, it also comes with security risks.
How does it work?
Built by Austrian developer Peter Steinberger, Moltbot is an open-source AI assistant that runs on individual computers (rather than in the cloud) and interacts with users through chats across a range of apps, including iMessage, WhatsApp, Telegram, Discord, Slack, and Signal.
Importantly, Moltbot can also monitor users’ calendars and other accounts and proactively send alerts, potentially leading to an important evolution in how AI systems are integrated into our daily lives. Meta is also reportedly experimenting with a chatbot that takes control by sending users an initial message, but this is clearly born out of the logic of engagement rather than practicality.
Related article: Stop it, Claude: Moonshot’s new AI model lets you create vibe codes from a single video upload
Rather than using its own large-scale language model, Moltbot leverages models from Anthropic and OpenAI. The assistant’s original name, Clawdbot, was a direct reference to Anthropic’s Claude chatbot, but Steinberger changed the name following a legal challenge from the company. (The new name suggests regrowth, since lobsters shed their skin, just as snakes shed their skin.)
Moltbot’s main appeal is that it combines the conversational abilities of Claude and ChatGPT with the ability to perform concrete actions within a user’s computer.
early feedback
Shortly after its release, Moltbot started making waves in the AI community. As of Wednesday afternoon, it already had 86,000 stars on GitHub, making it one of the fastest growing projects on the website to date. (Clawdbot was released on GitHub in late 2024, but the assistant’s viral outburst has occurred in the last few days.)
“I’ve been using @moltbot for a week and it feels like really early AGI,” one user posted on X on January 7th. “The gap between what you can imagine and what actually works has never been smaller.”
Also: Vibecode a Mac app in 8 hours using Claude Code, and it was more hard work than magic
Two weeks later, another user wrote that Moltbot felt like a huge paradigm shift for consumer AI. “When you experience @moltbot, you get the same shock you had when you first saw the power of ChatGPT, DeepSeek, and Claude Code. You know a fundamental shift is happening. [in] How to utilize AI. ”
The importance of silos
However, breathtaking early praise should not be taken as a guarantee of safety. On the other hand, if you decide to try your hand at Moltbot, you’ll need to be very careful, as it essentially requires you to hand over the keys to your account.
This problem creates a core tension for AI agents in general. The more autonomy AI agents have, the more vulnerable they are to facilitating injections and other cyberattacks. But in the case of Moltbot, the system’s ability to connect to a long list of messaging apps like WhatsApp means malicious actors have more potential entry points.
Related article: 10 ways AI could cause unprecedented damage in 2026
Many people, especially the 5×5 inch 2024 M4 Mac Mini (currently Available on Amazon for $499). Moltbot runs silently in the background and consumes very little power, making it ideal for always-on AI assistants. Even better, this approach means you don’t need to launch Moltbot on your personal or work computer, where all your passwords and other digital credentials are stored.
