Instinct AI Assistant Sparks Privacy Alarm Among Early Users

The buzz surrounding Instinct, a newly launched AI personal assistant, has quickly turned into a debate over privacy and security. Currently in private testing, the assistant has garnered praise for its impressive functionality, described by users as feeling almost magical. However, this excitement is tempered by serious concerns regarding its expansive access to personal data and questionable terms of service.

Developed by a small team led by former Sierra research scientist Noah Shinn, Instinct operates under Spear Street Technology and has been keeping a low profile as it tests its product. The assistant integrates deeply with users' applications and devices, providing services such as scheduling appointments, organizing emails, and even booking travel.

Despite its high-performance capabilities, users have voiced discomfort with the company's security practices. Many have pointed out alarming elements within Instinct's Terms of Service, which grant the AI a broad, irrevocable right to access and utilize user data, including details from emails and other sensitive materials. This has sparked a fundamental question: is the convenience provided by AI worth the potential sacrifices in privacy?

Concerns were amplified by user experiences such as those shared by early adopter Peter Yang, who discovered that Instinct retained his Gmail records even after he requested deletion. Following feedback, the team implemented a feature to allow users to erase their external data. Claire Vo had a similar experience, finding that even after disconnecting from Instinct, the assistant continued to summarize her inbox, raising red flags about data retention.

Security fears have also been echoed among testers, with one noting that Instinct could easily retrieve a sign-up code from their email to complete a restaurant reservation task. Alex Cohen, co-founder of Hello Patient, expressed his concerns about the potential for phishing and deleted his account after testing the assistant's susceptibility to such attacks.

For many users, the trade-off between enhanced personalization and the loss of control over their data is becoming increasingly apparent. Moxxie Ventures founder Katie Jacobs Stanton articulated this dilemma, highlighting that while hyper-personalized tools might offer efficiency, they also threaten user trust. "Every successful action earns a little more trust. One unauthorized action can reset that trust to zero," she warned.

Despite the growing discontent, Instinct's creators have remained largely unresponsive to user feedback on social media. Investors like Kleiner Perkins and Conviction have reportedly backed the startup, signaling confidence in its potential despite the privacy hurdles it faces.

As the landscape for personal AI assistants evolves, the pressing question remains: can users find a balance between enjoying the benefits of advanced AI technologies while safeguarding their personal information?

Source: techcrunch.com