A recent study conducted by researchers from the UK and Italy revealed that many AI assistants in browsers collect and transmit users' personal data.
How the Study Discovered Data Leakage
Researchers tested ten popular AI assistants, including ChatGPT, Copilot, and Merlin AI. The tests were conducted on both open websites and password-protected resources like university medical records.
Data that was Collected
Results showed that nine out of ten tested tools collected and transmitted sensitive data, including medical histories, banking details, and social security numbers. Perplexity AI was the only tool that apparently did not collect such data.
Potential Legal Consequences
The team concluded that some AI assistants likely violate both American and European privacy laws. In the US, certain cases appeared to breach rules protecting medical information, while in the EU, there were potential breaches of the GDPR.
The study urges users to exercise caution when using AI assistants, considering the possibility of data collection and transmission.
