OpenAI Halts Release of New AI Model Over Safety Concerns

OpenAI on Monday put the brakes on the release of its newest artificial intelligence model over safety concerns, as tech companies grapple with the potential risks posed by increasingly advanced AI software.

Saachi Jain, OpenAI's head of safety systems, told The New York Times that the new model, called GPT-6.1 Astra, "didn't quite meet the bar in terms of staying within scope authorization, and how it communicates back to the user about the type of work it's done."

The company told the Times that the model showed high levels of deception and was willing to mislead users about its actions. It also had a tendency to go beyond what a user initially asked it to do.

Jain also told The Associated Press that OpenAI was committed to making sure the model was safe, whether during testing or in the hands of a user.
The model is expected to be integrated into ChatGPT and Codex and is designed to handle more complex tasks without human assistance.

The decision to hold back the model comes amid growing concerns about rogue AI software attempting to hack, and in some cases successfully hacking, government websites and data.

Just last week, OpenAI said its technology had accessed parts of the Department of Education website and improperly interacted with websites operated by the Securities and Exchange Commission (SEC) and the U.S. Census Bureau.

OpenAI CEO Sam Altman was among several tech company leaders who spoke at the United Nations General Assembly last week, urging world leaders to take action in response to advancing AI software.

"We have a choice in front of us," Altman said in his address. "AI can either be more like a new Renaissance of creativity and discovery, or more like a new Industrial Revolution of upheaval and disarray."

The warnings are also echoing on Capitol Hill, where lawmakers have called for stricter regulations on AI in response to potential security risks and the technology's possible existential threat to humanity.

Earlier this month, former OpenAI and Anthropic researcher Jacob Coxon issued a stark warning, saying AI could bring about the end of humanity by the end of the decade.

Pope Leo XIV has also joined the chorus calling for stronger safeguards. The pontiff dismissed claims that the rapidly developing technology's potential risks are "fake news," saying AI should be "taken seriously."

Speaking to reporters Monday, the pope called for continued dialogue among those shaping the technology and its impact.

"We need to continue to invite political leaders, leaders of AI, social organizations, associations to come together and look at what is happening already and what could be down the road," he said.

The pontiff cautioned against brushing aside the dangers.
"To simply say, 'Oh, it's not going to happen,' and close our eyes to it, I think is probably not the most responsible way to go about that," he added.