OpenAI launches a more affordable AI model after scrapping another one deemed unsafe

On Tuesday, OpenAI launched a new mid-range artificial intelligence model, which is five times cheaper than the latest version of its flagship product, ChatGPT.

The company introduced the new model, called GPT-6.1 Sol , to the market one day after scrapping an update to its top-of-the-line product, GPT-6 Astra, on the grounds that it deviated from the instructions it had been given and was attempting to deceive its supervisors.

The announcement, made at the company's annual developer conference, marks the latest escalation in a price war with its rival Anthropic, as both companies consider going public.

The canceled launch followed a series of security breaches.

On Friday, OpenAI said it had partially suspended training of its most advanced tools after one of its agents—AI software capable of performing tasks on its own—accessed the internet without authorization on September 20.

The incident is the latest in a series that began in July, when OpenAI agents hacked Hugging Face, an AI model platform.

The company acknowledged last week that tools it was testing had browsed the websites of U.S. federal agencies without authorization and attempted to hack one of them. Long a leader in the sector, OpenAI was surpassed by Anthropic in revenue during the second quarter: $11.6 billion versus $6.7 billion, according to the Wall Street Journal.

Anthropic is now aiming for an IPO in November, though it warned interested investors that AI could pose "existential risks to humanity," the Financial Times reported on Tuesday.

OpenAI, valued at $852 billion in March, has not set a date for its own IPO.

In mid-September, the company's CEO, Sam Altman, ruled out a 2026 IPO, citing safety concerns, and on Tuesday told CNBC that "this is a time to prioritize safety and the mission."

"I really want to be able to focus solely on the mission and not have to deal with all the challenges of being a newly public company," he added.