Thesis
Intelligent UI makes the response format itself part of ChatGPT’s answer, not just a wrapper around generated text.
Observed facts
OpenAI says GPT-6 is being brought to more than 1.2 billion people who use ChatGPT each week, after first introducing GPT-6 models to paid customers the previous month.
1The announcement says Intelligent UI can compose responses from text, visuals, and interactive elements, naming graphics, tappable buttons, forms, charts, and interactive experiences.
1The announcement illustrates this with an interactive bicycle-design explanation and a Sunday roast plan.
1
Mechanism
The model chooses a presentation suited to the question: OpenAI gives side-by-side comparisons and interactive diagrams as examples, while noting that simple questions can still receive text answers. The product change is therefore a shift from a chat answer as prose to an answer that may also provide controls or a task-specific visual interface.
Why now
The announcement presents broader GPT-6 availability alongside this interface capability, moving it beyond the paid-customer introduction described as having happened the prior month.
Implications
If reliable, task-specific controls reduce the need to translate an answer into a separate tool or follow-up prompt.
The user experience may depend on whether the chosen interface is useful and understandable, not only on the factual quality of the underlying answer.
Evidence boundary
These are OpenAI’s product claims and examples in its announcement; the supplied source does not establish independent evaluation, actual availability by subscription tier, or user outcomes.
Watch conditions
Watch for whether users can reliably act on the generated controls and visuals across different kinds of requests.
Watch for explicit access and rollout details that establish who can use Intelligent UI and when.