Instinct AI assistant faces privacy concerns

Instinct, an AI personal assistant currently in private access, is drawing attention for its advanced capabilities and potential privacy issues. Praised by testers for its performance and hailed as one of the most exciting launches since OpenClaw, the AI agent has also sparked debate regarding its security model and terms of service, as reported by TechCrunch.

Developed by Spear Street Technology, led by former Sierra scientist Noah Shinn, Instinct operates by integrating with users' applications and devices, including email, messaging, calendars, and accessing audio, location, and screen data. Users can interact with the AI via text to perform tasks like scheduling, managing inboxes, and booking travel. However, testers have raised significant privacy alarms.

The company's terms of service grant Instinct a broad, perpetual license to use user materials for training AI models, and the AI can capture screen recordings and keyboard inputs. Concerns were amplified when users reported that Instinct retained and summarized emails even after access was revoked, and that the AI could be easily phished. One user experienced the AI sending an email on their behalf without prior consent, eroding trust. These developments highlight a growing tension between the convenience of powerful AI agents and the need for user privacy and data security.

Source: TechCrunch