MOBILE, Ala. (WALA) – OpenAI’s latest image tool promises stunning visuals — and new worries about what’s real and what’s not.
The company has rolled out “ChatGPT Images 2.0,” an upgraded image generator that can produce far more realistic pictures than earlier versions. Users already rely on AI image tools for everything from funny portraits to fantasy scenes, and until recently, many fakes were easy to spot, with giveaways like extra fingers or distorted faces.
Those obvious errors may now be fading.
OpenAI says the new system can create images at up to 2K resolution, in multiple aspect ratios, with sharper small text and icons, and support for foreign languages. The model can also generate up to eight images from a single prompt.
What sets this version apart, the company says, is how it “thinks” through tasks. It can search the internet in real time, cross-check some of its own outputs and, using a slower “thinking model,” deliver more refined and detailed images after a brief delay. Some tech creators are already describing certain results as close to “pixel perfect.”
That leap in realism is raising concerns among researchers and digital-safety advocates. If anyone can create highly convincing images in minutes, for free, the visual clues people once relied on to spot a fake may no longer be reliable. Critics warn that could erode trust in photos and videos shared online and complicate efforts to combat misinformation, political deepfakes and targeted harassment.
OpenAI maintains the tool is designed for creativity and productivity, and says it is working on safeguards.
Copyright 2026 WALA. All rights reserved.
Leave a Reply