ChatGPT has entered the group chat. Texters have…feelings about it.

OpenAI launched a plugin for Apple Messages on Thursday that lets ChatGPT on a Mac search text messages, draft replies, and send them, sparking a debate about AI's role in personal communication. The feature, available in Codex and ChatGPT Work on desktop, is off by default and only reads messages when prompted, but some users have threatened to cut off contact with anyone who uses it. OpenAI's head of developer experience, Romain Huet, suggested prompts like asking who you talk to most or what you forgot to follow up on.

A hot new bombshell has entered the group chat: ChatGPT. OpenAI announced a new plugin for Apple Messages https://www.businessinsider.com/iphone-apple-text-message-imessage-recording-voice-memo-be-careful-2025-7 on Thursday. On a Mac, ChatGPT can now search your text messages, draft replies, and hit send. The company suggested using prompts to look for birthday or spam messages to delete. The launch cracked open a debate: how human should your text chats be? Some made it clear that they considered AI texts crossing a line. "If I find out that AI has been texting me on your behalf, I will never speak to you again," one X user https://x.com/BigImpactHumans/status/2090568331317108786?s=20 wrote. The OpenAI texting launch OpenAI's new launch is available in Codex and ChatGPT Work on desktop. Romain Huet, OpenAI's head of developer experience, gave some use cases https://x.com/romainhuet/status/2090587684272021760 : "Ask it who you talk to most, what you talk about, or what you forgot to follow up on " The company has been expanding its slate of features — and the data it requests access to. Last week, the company rolled out Computer History, which lets ChatGPT learn from the actions you take on some apps and websites. Like the messages feature, Computer History is off by default. i will buy a pirate ship and force my friends to physically walk the plank if they hand the groupchat over to OpenAI — tom zehentner @tomzehentner https://t.co/3uiYp6gJFP August 20, 2026 Ironically, the messaging feature looks a lot like Apple's new Siri update, which is slowly rolling out. In July, Apple filed a trade-secrets lawsuit https://www.businessinsider.com/apple-sues-openai-trade-secret-theft-2026-7 against OpenAI over the development of its hardware device. One wonders how the company feels about this new feature. As for texting, let's face it: the group chat has already grown artificial https://www.businessinsider.com/rise-chatgpt-groupchats-texting-openai-friendship-marriage-loneliness-generative-ai-2025-7 . People are drafting their messages with chatbots and adding their OpenClaw agents to the chat. The OpenAI launch is a ramp-up of what's already been in motion. And yet, as emails become more AI-ified, texting has largely remained a communication channel with undeniable humanity. Reddit cofounder Alexis Ohanian https://www.businessinsider.com/alexis-ohanian-much-of-the-internet-is-now-dead-2025-10 said as much: "I think we'll see a next generation of social media emerge that's verifiably human, because it's all going down in the group chats now." Fighting for the group chat The launch immediately caught backlash. One user https://x.com/Midnight Sweaty/status/2090761745287798824?s=20 said that he was "never going to talk to you again" if you used the feature. Another said https://x.com/ cowbutch/status/2090666390516633835?s=20 she'd block your contact. "If anyone AI emails or texts me I hope you know youre effectively dead to me," one X user wrote. "I dont have the energy or time for a relationship with someone who invests no effort." Danisha Carter https://x.com/danishacarterr/status/2090635193790967996?s=20 called the feature an "open-book 'are you an idiot' test." Not everyone was feeling so gloomy. Creative Strategies analyst Max Weinbach https://x.com/mweinbach/status/2090509815290167589 wrote that it was a "game changer" with the amount of work he does over text. Yana Welinder wrote that she will "actually be able to find things in the black hole of iMessage now." texts from my AI-pilled friends now that chatGPT has iMessage integration — brett goldstein @thatguybg pic.twitter.com/DT8DbAYsje August 20, 2026 There's also the privacy question. The feature must be explicitly installed and enabled, an OpenAI representative said, and does not index on users' messages. ChatGPT will only read a text message when the user prompts, they said. Still, users seemed on edge. "Is nobody else semi-weary about giving ChatGPT unfettered access to EVERYTHING?" Stephen Diedrich https://x.com/stephdiedrich/status/2090565281789665590?s=20 asked. Tom Goodwin https://x.com/tomfgoodwin/status/2090666657760915892?s=20 said that he generally has a "very very relaxed sense of security," but that using the feature seemed "incredibly naive." "I hope all my messages about how much I hate ChatGPT hurt its feelings," wrote Jacqueline Smith https://x.com/JackieSmith114/status/2090637842342875315?s=20 . Are AI-generated texts fair game or a turn-off? Let us know your view in the comments below: