Microsoft to Redeploy AI Tool: Privacy Concerns Met with System Improvements

Microsoft has announced that it'll relaunch its AI-powered tool, Recall, with updated privacy features in November. The tech giant has taken feedback into account after the initial unveiling in May 2024 sparked concerns over privacy issues. 

The new version will now require user consent to function, a change from its initial design, which had it on by default. Recall is designed to help users access past activities on their PC through artificial intelligence, including files, photos, emails, or browsing history. It works by taking regular desktop screenshots, which fuelled worries over the amount of sensitive data it would process, prompting the UK's data watchdog, ICO, to investigate. 

Changes to Recall have reassured the ICO, but they will continue to monitor development. New security measures announced include the encryption of snapshots and associated information, biometric login for screenshot access, and exclusion of sensitive information such as credit card details by default. 

Addressing security and privacy concerns, Pavan Davuluri, Microsoft's corporate vice president of Windows and devices, explained that all snapshots are encrypted and the system offers tools to control privacy. He assured that screenshots can only be accessed through biometric login, and sensitive information wouldn't be captured by default.

Despite the reassurances, Professor Alan Woodward, a cybersecurity expert, urged caution, advising users to wait until the tool has been thoroughly tested in real-world environments.

- CyberBeat