Microsoft has been testing a new AI feature, Recall, which records user activities on Windows Copilot+ PCs and helps them find anything from the past.
Microsoft has officially declared the test launch of its most controversial and recent AI introduction, the Recall AI search feature. The search function of Recall does not look up the internet to provide results; instead, it dwells on past activities on Windows Copilot Plus PCs to help find anything from any time. AI Recall will be released to its insider program members beginning in October.
″With a commitment to delivering a trustworthy and secure Recall (preview) experience on Copilot+ PCs for customers, we’re sharing an update that Recall will be available to Windows Insiders starting in October. As previously shared on June 13, we have adjusted our release approach to leverage the valuable expertise of our Windows Insider community prior to making Recall available for all Copilot+ PCs.” Microsoft updated its original June blog on Wednesday.
The Recall function does not have a controversial capability, but the process it has to go through can create high-risk security vulnerabilities. Researchers have indicated that the screenshots it takes of user activities can store critical information like passwords, explicit media, and financial data that hackers can abuse. Though the Recall AI feature will first launch in Microsoft Copilot Plus PCs, it may extend to Windows devices with similar hardware capabilities.
Originally, Microsoft said that the AI Recall feature would be enabled by default, but it has now changed that to being turned off by default. Users will have to opt-in to enable the Recall feature, and only then will the AI tool begin recording and monitoring the activities. Microsoft has also assured that it will launch the new AI feature with added layers of security to minimize the risks.
AI Recall takes screenshots of everything and is completely an on-device software. Microsoft’s close rival, Apple, has also launched local AI tools that do not fully require online connectivity. It is evident that the tech giants are now in a new race to provide localized AI tools and capabilities to every user that remains private.
The danger here is in the devices’ security access layers, which sophisticated hackers and bad actors can breach. Once they are in the device, the AI or the security protocols do not usually detect malicious activities posing multi-faceted threats. The skepticism around these controversial statements revolving online globally will only be answered when the artificial intelligence feature completes its first insider test phase.
