The British Telegraph reports that thousands of people have downloaded a WhatsApp personal assistant designed to book restaurants, handle business emails, and manage children’s homework. The AI-based personal assistant, called Clawdbot, has gone viral in recent days, with users claiming it represents the next major step in the use of artificial intelligence, following the growing popularity of programs such as ChatGPT.
However, it has also raised serious security concerns, with experts warning that granting the bot access to private messages and even personal finances could become “a disaster in the making”.
Clawdbot, created by London-based technology entrepreneur Peter Steinberger, operates as an “agent”, meaning it can act independently to carry out tasks, browse websites, and control applications and email inboxes. The nature of Clawdbot makes it more powerful than systems such as ChatGPT, which are primarily known for answering questions or generating images.
Users can control the bot by connecting it to a messaging app such as WhatsApp, allowing them to send commands directly from their phone. The bot updates over time in order to learn users’ habits and preferences.
Matt Van Horn, a technology entrepreneur, developed a way to connect the bot to his Tesla vehicle, allowing him to send a message to Clawdbot to remotely defrost the car. John Oringer, an American technology billionaire, said: “Clawdbot is insane… it’s what we all needed”.
Although the number of Clawdbot users has not been disclosed, more than 15,000 people have joined an online chat room dedicated to the bot. More than 36,000 people have “liked” the project on GitHub since its release, up from 7,000 Recently. Alex Finn, a software developer, said he sent a message to Clawdbot to book a table at a restaurant in Silicon Valley.
When the bot failed to complete the booking through the website, it used ElevenLabs’ voice tool to call the restaurant and reserve the table.
Dan Paguin, a technology entrepreneur, said he used the app to read emails from his child’s school and summarize key points such as exam dates. “Clawdbot drove some people in my child’s parent group crazy yesterday,” he said. Other users noted that they used the bot to manage finances and schedules.
Dave Morin, a venture capital investor, said: “This is the first time I’ve felt like I’m living in the future since the launch of ChatGPT”. Another user said it was “Siri done right”, referring to Apple’s voice assistant, which has become less popular in recent years.
The high level of interest in the bot has led to increased demand for Apple’s Mac Mini computers, which many people use to run the program. Searches for Mac Mini jumped recently, according to Google data.
However, security experts have warned against installing the program in a way that could grant it access to sensitive data. To operate, AI agents require deep access to computer systems, enabling them to read private messages, handle files, and write code, which could lead to the deletion of important documents or the installation of viruses on a computer.
“This is a disaster on the way,” said Alan Woodward, a computer security expert at the University of Surrey. “Give artificial intelligence the ability to speak and act on your behalf, in addition to access to all your private data. What could possibly go wrong?” Woodward added that the system is “far beyond being suitable for non-expert users”.
Some people who tried to set up the bot discovered that it sent messages to people in their contact list in what appeared to be a random manner. Shares of Cloudflare, a website security company used by many to protect Clawdbot systems, rose by 10% recently. Steinberger, the Austrian developer who created the app, has not yet made any money from it.