OpenAI unveils 'dots' AI assistant as safety concerns delay next model

OpenAI unveiled a new AI assistant called "dots" at its annual developer event in San Francisco, even as concerns over safety have delayed the release of a more advanced model, according to the BBC.

Chief executive Sam Altman spoke at the event, which OpenAI uses each year to showcase new products and features to developers building on its technology. The unveiling of dots came alongside acknowledgment that a separate, more capable model has faced delays tied to safety worries.

Balancing new products against safety concerns

The dual announcement illustrates the tension AI companies increasingly face between shipping new consumer-facing products and addressing safety questions that can slow the release of more powerful underlying models. OpenAI's decision to delay the more advanced model while still introducing dots suggests the company is trying to keep up its pace of product releases without abandoning caution around its most capable systems.

The BBC's report did not detail the specific capabilities of dots or the precise safety issues that have held back the delayed model, though the framing suggests the two are viewed by OpenAI as separate tracks of development moving at different speeds.

OpenAI's developer events have historically served as a venue for signaling the company's product roadmap to the wider tech industry, and this year's event continued that pattern by pairing a new release with a candid acknowledgment of an ongoing delay elsewhere in its lineup.

The announcement comes amid continued scrutiny of AI safety practices across the industry, with companies facing pressure from regulators, researchers and the public to demonstrate that new systems are being tested and released responsibly.