ChatGPT Adds Photo Stickers, Chat Lock, Mac Messages Access

A Flurry of New ChatGPT Features Signals a Bigger Ambition

OpenAI is rapidly expanding what ChatGPT can do beyond text generation, rolling out a mix of playful and privacy-sensitive tools that push the assistant deeper into everyday messaging and device control. The most visible addition lets users turn personal photos into custom sticker packs for use in iMessage and WhatsApp, giving ChatGPT a lighthearted, consumer-friendly upgrade that fits neatly into how people already communicate with friends and family 1.

Locking Down Sensitive Conversations

While the sticker feature is aimed at fun and shareability, other discoveries point to OpenAI addressing growing concerns about privacy. Code buried in the latest Android version of the ChatGPT app suggests the company is preparing a way to lock individual chats behind a PIN or fingerprint scan 23. This would let users shield sensitive conversations — medical questions, financial details, or personal journaling — from anyone else who might pick up their phone. The feature has not been officially announced, but its presence in inactive code is typically read as a sign that a public rollout is being tested internally before wider release 23.

Reaching Into Messages and Mac Systems

The most striking development involves ChatGPT's expanding access to users' actual messaging apps. A new plugin allows ChatGPT to search through Apple Messages on a Mac and, in some cases, send replies on the user's behalf 4. One reporter testing the feature described an unsettling experience in which granting ChatGPT access to Messages effectively gave it a level of control over the Mac that went further than expected 4. A related report frames this capability more bluntly, describing an AI that can read iMessages and respond without explicit permission each time, raising questions about consent and oversight when an assistant is empowered to act autonomously inside a user's private conversations 5.

Why It Matters

Taken together, these updates illustrate a broader strategy: OpenAI is trying to make ChatGPT feel less like a standalone chatbot and more like an embedded layer across a user's digital life — capable of creating shareable content, protecting private data, and even managing communications directly. That trajectory mirrors trends across the smartphone AI assistant space, where companies are racing to integrate generative tools into browsers, messaging platforms, and operating systems.

But the coverage also highlights the tension inherent in that ambition. Features like sticker generation are low-risk and likely to be popular. Chat locking responds directly to user demand for better privacy controls as AI assistants store increasingly sensitive information 23. Yet the ability for ChatGPT to read and reply to personal messages without clear, repeated user consent raises real concerns about autonomy, security, and unintended actions taken on a person's behalf 45. As AI assistants gain more system-level access, the balance between convenience and control will likely remain a central point of scrutiny.