Urgent.News

What's breaking now, across thousands of outlets.

AI

ChatGPT Can Take Over Your iMessages. But Do You Really Want That?

OpenAI has a new feature that allows its AI chatbot to read your texts and write new ones.

ChatGPT Can Take Over Your iMessages. But Do You Really Want That?

OpenAI has recently introduced an AI feature for ChatGPT that could potentially scan and interact with Apple's iMessage service. This new capability, known as the Apple Messages Plugin, is currently only accessible to ChatGPT Work and Codex users on Mac desktops. The plugin allows users to read, write, and send messages on their behalf, but Apple Messages Plugin is not yet available for mobile devices.

Before activating this feature, users must go through a series of permission procedures. OpenAI assures that it won't store users' messages, and the plugin will only run locally on the user's Mac computer, not on OpenAI's cloud servers. However, despite these assurances, the feature could raise privacy concerns for Apple, a company that has long marketed itself as a strong advocate for customer privacy.

Apple and OpenAI had previously partnered in June 2024 to integrate ChatGPT into iOS, iPadOS, and MacOS systems. Yet, their relationship took a turn in July 2024 when Apple sued OpenAI, alleging the theft of trade secrets and intellectual property. OpenAI has been embroiled in numerous lawsuits over the past few years, with accusations ranging from misusing customer data to sharing it with other companies.

Adding the new ChatGPT feature involves navigating through the ChatGPT Plugins menu, selecting Apple Messages, and installing the plugin. A permissions screen will appear, notifying users that their message history on Mac will be accessed by the Apple Messages app. To use this feature, users must change their privacy preferences in System Settings on their Mac, including enabling Full Disk Access. This will grant ChatGPT permission to access contact names and automation tools.

Once the Apple Messages plugin is installed, users can activate it by typing '@' followed by the plugin name or using the '+' menu within ChatGPT. With this plugin, users can search messages, catch up on conversations, draft and send replies, and more. OpenAI showcased a video demonstrating the feature, illustrating how users can ask ChatGPT to review messages from the previous day to identify any follow-ups required.

While the potential capabilities of this feature are intriguing, the privacy implications cannot be ignored. In light of OpenAI's history of legal challenges related to data usage, it remains to be seen whether Apple users genuinely desire their messaging app to be accessible to OpenAI's AI chatbot.

Written by urgent.news from CNET's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at cnet.com →

More in AI

A modular offline multilingual speech-to-speech translation system with confidence-aware ASR routing and language-aware TTS

Scientific Reports, Published online: 22 August 2026; doi:10.1038/s41598-026-67788-0 A modular offline multilingual speech-to-speech translation system with confidence-aware ASR routing and…

  • Modular offline system integrates language identification, routing, translation, and synthesis
  • Routing threshold of 0.70 lowers recognition errors in Indic languages
  • Mean model-only CPU latency of 25.41 seconds for offline multilingual speech translation

Multi-label arrhythmia classification from Holter scatter plots using deep learning and clinically oriented thresholds

Scientific Reports, Published online: 22 August 2026; doi:10.1038/s41598-026-68106-4 Multi-label arrhythmia classification from Holter scatter plots using deep learning and clinically oriented…

  • Clinically oriented thresholds improved specificity to 0.885 but decreased macro-F1 score to 0.693.
  • Scatter plot-based deep learning could serve as efficient preliminary screening for Holter analysis.

Machine learning incorporating trabecular bone score improves lumbar fracture risk stratification in type 2 diabetes

Scientific Reports, Published online: 22 August 2026; doi:10.1038/s41598-026-67932-w Machine learning incorporating trabecular bone score improves lumbar fracture risk stratification in type 2…

  • Machine learning model using trabecular bone score improves lumbar fracture risk prediction
  • Study of 261 T2DM patients over 50 found significant improvement in fracture risk prediction
  • Logistic regression algorithm achieved highest performance with AUC of 0.818

Publishers Shift Strategy in AI Era

Japanese online publishing platform note is using artificial intelligence to review its management meetings and analyze company-wide activity as it prepares for a future in which AI increasingly…

  • Japanese online platform Note adopts AI to manage operations amid digital shift.
  • Emphasize closed vs accessible data for AI, consider paywalls.
  • Transition from SEO to AIO, GEO, LLMO optimization for AI-generated answers.

More from Friday 21 August →