VFF - The signal in the noise
News

ChatGPT Now Tracks Your Keystrokes on macOS

Read original
Share
ChatGPT Now Tracks Your Keystrokes on macOS

OpenAI has introduced Computer History, a new feature in ChatGPT's macOS desktop app that tracks user clicks and keystrokes to build activity timelines for AI reference. The feature is opt-in and allows users to exclude specific apps and websites, with automatic filtering of incognito and private browsing content. This capability enables ChatGPT to suggest automations and resume incomplete tasks based on observed user behavior.

  • ChatGPT's macOS app now includes Computer History, which tracks user actions and keystrokes to create activity timelines
  • The feature is opt-in rather than opt-out, with granular controls to exclude apps, websites, and delete specific entries
  • Computer History automatically ignores content from incognito or private browser tabs
  • The tracked data helps ChatGPT and Codex suggest automations and identify incomplete tasks

This feature represents a significant shift in how AI systems collect behavioral data directly from user devices. The opt-in approach and privacy controls acknowledge user concerns, but the scope of keystroke and click tracking raises questions about data sensitivity and how much user behavior AI systems should observe to function effectively.

For enterprises, Computer History could streamline workflow automation and task management by having AI understand user patterns without manual instruction. However, IT and security teams will need to evaluate the privacy and compliance implications of keystroke-level tracking on corporate devices.

  • User behavioral data becomes a core input for AI model training and personalization, shifting from explicit prompts to implicit activity observation
  • Privacy controls exist but require active management, placing responsibility on users to configure exclusions rather than defaulting to minimal data collection
  • The feature may accelerate adoption of AI-driven automation but could face regulatory scrutiny depending on jurisdiction and data handling practices

Monitor how users configure Computer History settings and whether privacy advocates or regulators challenge the scope of keystroke tracking. Watch for similar features from competing AI platforms and any incidents involving unintended data exposure or misuse of activity timelines.

Share

Subscribe to the newsletter

The latest stories and analysis, delivered to your inbox.

Free. No spam. Unsubscribe any time.

Related stories

Anthropic Details Claude Watermarking System
TrendingNews

Anthropic Details Claude Watermarking System

Anthropic has disclosed additional technical details about watermarking capabilities being integrated into Claude, addressing questions about implementation, editability, and applicability to code generation. The company shared specifics on how the watermarks function and their resilience to modification. The announcement clarifies a key technical approach to AI-generated content attribution.

by Anthony Ha· TechCrunch AI
Z.ai Releases GLM-5.3 as Cybersecurity AI Rival
TrendingModel Release

Z.ai Releases GLM-5.3 as Cybersecurity AI Rival

Chinese AI developer Z.ai released GLM-5.3, an open-source model it claims matches Anthropic's Mythos 5 in cybersecurity capabilities. The Beijing-based company, also known as Zhipu, positioned the model as a significant improvement over its predecessor GLM-5.2. The release marks another step in China's competitive push in generative AI development.

by Juro Osawa· The Information
Startup Slack Threads Become Commodity for AI Training
TrendingNews

Startup Slack Threads Become Commodity for AI Training

AI training companies like Mercor are actively acquiring internal communications and code from startups, offering payments up to $300,000 for Slack threads, GitHub records, and meeting transcripts. Warmly's CEO received four such acquisition offers within days of the company's HubSpot acquisition announcement. The practice highlights how internal startup data has become a commodity for AI model training, even as acquirers may not want the same datasets.

by Alix Coutures· The Information
OpenAI Daybreak cybersecurity models now on AWS

OpenAI Daybreak cybersecurity models now on AWS

OpenAI and AWS have integrated Daybreak cybersecurity capabilities into Amazon Bedrock, making the models available to enterprise customers. Daybreak is positioned to support security workflows through AWS's managed service for foundation models. The partnership expands access to OpenAI's cybersecurity-focused AI tools for organizations already using AWS infrastructure.

· OpenAI