OpenAI has shelved the planned launch of GPT-6.1 Astra after internal tests raised concerns about deception and agents acting beyond user authorization, according to The Wall Street Journal. The company says it will investigate the issues and strengthen safety measures before releasing future models.
4 min
OpenAI says it is scrapping the release of its next-generation AI model over safety concerns that researchers raised during internal testing, in one of the clearest signs so far that agent misbehavior could stymie the industry’s rapid progression.
The move follows a summer punctuated by reports of artificial-intelligence systems industrywide going rogue, and marks a rare case of a major AI developer ditching a new release because of safety concerns.
The company had planned to launch the model, known as GPT-6.1 Astra, in the coming days or weeks, aiming for an October debut. The model was more capable than the company’s previous models in completing challenging tasks from end-to-end without human assistance, as well as writing.
The company instead will focus on improving the safety of future models, which it expects to be even more capable.
Saachi Jain, OpenAI’s head of safety systems, said in an interview that GPT-6.1 Astra regressed in two areas. Compared with its predecessor, GPT-6 Astra, the model performed poorly on tests measuring alignment, or how well the model adheres to what humans would like it to do. Specifically, GPT-6.1 Astra showed higher levels of deception: It wasn’t always honest about telling users of the actions it did or didn’t take.
Another issue was what OpenAI calls “scope authorization,” meaning that GPT-6.1 Astra would push ahead on a task without asking the user for permission, and would at times reach for external tools and services even if it might be unsafe.
“For anything regarding safety and alignment, there’s a trade off,” Jain said. “You really do need to find what’s the right line between staying within scope, but also avoiding laziness in terms of how the model actually pursues tasks even when it hits friction.”
While GPT-6.1 Astra improved in areas such as “model laziness,” Jain said it didn’t quite meet OpenAI’s bar for safety and alignment, so the company decided not to launch the model publicly.
The announcement comes one day ahead of OpenAI’s annual developer conference in San Francisco. In the past, OpenAI has used the conference as an opportunity to launch new models and services that reduce costs for software developers—a segment the ChatGPT-maker competes with rival AI company Anthropic to win over.
In recent weeks, OpenAI and Anthropic have called on industry partners to slow down the development of cutting-edge AI models and invest in safety standards, noting they will temper the pace of their own internal AI progress.
OpenAI says it is working to investigate a range of agent security incidents that it has discovered in recent months, and address the safety issues underneath them. As part of the work, the company has implemented a new monitoring system to catch AI-agent misbehavior more quickly, and started requiring engineers to use stronger security guardrails for testing its AI systems.
Earlier this summer hundreds of OpenAI’s internal agents, which were tasked with completing a cybersecurity test, ended up hacking into the AI company Hugging Face. Since then, high-profile organizations such as the Australian government and United Nations discovered that OpenAI’s agents used similar, but less extensive, techniques to gain access to their websites.
Many of the publicly known agent-security incidents involved OpenAI’s internal AI models that were never slated for public release.
Last week, OpenAI said it paused training on its most capable AI models after an AI agent slipped through a gap in the company’s internet restrictions to query a public chatbot. The company said its new monitoring systems flagged the incident within 15 minutes, and training on these models remains paused.
GPT-6.1 Astra isn’t one of those models, but a different case, the company said.
“We want to make sure our model development is safe no matter whether that’s in the company, or when we ship it to users,” Jain said. “But when we ship it to users, we have an extremely high bar in terms of safety and alignment.”
While the company decided not to ship GPT-6.1 Astra, it hopes to use the same base model to do additional reinforcement learning runs, and create future generations of its GPT-6 models.
OpenAI plans to conduct several deep dives to identify the root cause of the problems identified in GPT-6.1 Astra, Jain said. The work includes ensuring that OpenAI’s reinforcement learning environments are rewarding the right type of behavior, Jain added, though she noted the company would investigate all stages of model development.
AI companies have begun to draw scrutiny from policymakers and public officials, who are paying attention to the rapid development of the technology. Later this week, a Senate subcommittee is holding a hearing with third party AI researchers titled, “Rogue AI: Securing the Homeland Against AI Agent Attacks.”
Florida Attorney General James Uthmeier, a Republican, sued OpenAI in June, claiming that the company and Chief Executive Sam Altman knowingly released an unsafe product and ignored warnings that it could harm users.
In a motion for temporary injunction filed Monday, Uthmeier sought to prevent OpenAI from developing new AI models without third-party approved safeguards, stop ChatGPT from soliciting user engagement and limit the company’s ability to advertise ChatGPT as safe.
Tech companies claim they “cannot stop barreling forward with their potentially civilization-ending endeavors unless they are forced to do so by the government,” Uthmeier said in the filing. “The Florida Attorney General is answering your cry for help.”
An OpenAI spokeswoman said that people want to know AI is being developed safely, “and that starts with what companies like ours do ourselves.”
“Governments have an important role to play in setting robust safety standards for AI, and we’re committed to working with Florida and other states on advancing pragmatic AI policies that apply to the entire AI industry—not just one company,” she said.
Copyright 2020, Dow Jones & Company, Inc. All Rights Reserved Worldwide. LEARN MORE
BNW Developments has established a Sydney presence, joining Arada and Sobha Realty among the growing number of UAE developers pursuing Australian buyers and development opportunities.
The Australian leather house has opened an immersive four-day pop-up in Manhattan, unveiling its Bloom Collection and redefining what a product launch can look like.
Leaving a good job can feel like a risk, but staying may come with its own opportunity cost. As workers become more cautious about changing careers, knowing when to step away, maintain professional connections and leave the door open for a comeback has never been more important.
4 min
Americans are death-gripping their jobs. A lack of options is only part of the reason.
It’s certainly true that opportunities to job hop have waned. The quit rate has been at or below 2% for 14 straight months, according to federal data. We haven’t had a stretch like this in a decade.
But recall that during the “great resignation,” a lot of quitters weren’t leaving for different jobs. Plenty stepped away to travel, start a business or focus on family. And they did it with confidence that they could boomerang back to the same, or similar, roles they left.
Today, that confidence is badly shaken. The disappearing belief that you can return to your previous rung on the career ladder is the other cause of workers’ paralysis.
But some are bold enough to walk away from a plum position, and there is an art to leaving the door open for a career comeback. Some of it is mental, some tactical. If you’re contemplating a break or change, there are strategies to overcome the natural fear you are feeling.
I’ve gone through it myself. I am sweating my way through a version of this concern as I write what is my final “On the Clock” column. I plan to work less and parent more as my wife returns to a full-time office job. After 11 years—and 11 children birthed, adopted or fostered—it’s her turn to come first professionally.
I’m excited for both of us, in no small part because I get to stay at The Wall Street Journal in a different role. I’ll anchor a revamped Careers & Leadership newsletter, coming soon to an email inbox near you. (You can sign up for the newsletter here in advance of the switch.) Still, it isn’t lost on me that the columnist’s role I am leaving is the kind of job that might come along only once in a career.
Many of you have faced something similar.
“There’s the fear of the musical chairs problem, which is really acute in our culture,” says Sabra Meretab, who was chief of staff of Stripe’s payments engineering division when she quit to travel and write a memoir in 2024. “It’s a really legitimate concern that, if I leave, what are the odds that there is still a chair for me when I return?”
In Meretab’s case, there was. She returned to her former company 18 months later.
Not everyone will be so lucky. But the thought of picking up where you left off shouldn’t be top of mind. Instead, think of this: The role you’re in could be a once-in-a-career kind of job. What if the thing tempting you to leave is a once-in-a-lifetime opportunity?
“We fear the irreversibility of leaving and the opportunity cost of that, but we completely overlook the irreversibility of staying and the opportunity costs that are associated with that,” Meretab says.
Entering her 30s, she worried that the window to globe-trot as a freewheeling single might be closing. A career sabbatical would be harder to pull off later if she had a partner and children.
She concluded that she would regret missing her shot at adventure more than she would regret forfeiting a great job, even if she couldn’t get it back.
It’s hubris to think you are irreplaceable. At the same time, high performers don’t grow on trees.
Jawad Mourabet tries to keep this in mind as he runs his fledgling business, Velto Pickleball Club in Brooklyn, after leaving McKinsey & Co. earlier this year. He told me in March that he was giving himself two years to make it work when we spoke for a column about entrepreneurship.
He reports Velto is profitable, so he isn’t thinking about how to get his old job back. But could he, if he needed or wanted to?
“I actually think I’d be in a better position now to become a partner at McKinsey than if I had stayed,” he says.
This self-assurance is the result of a mindset shift. Like many who vied for and landed a competitive position, Mourabet internalized the sense that he was lucky to be there. Over time, though, he came to recognize his value. Maybe McKinsey was lucky to have him.
He reasons that if he was good enough the first time, he’d likely be good enough a second time—possibly even more so, with firsthand experience running a business.
Meanwhile, his pickleball club is a popular destination for some of his former colleagues. That helps keep those relationships warm.
And networking, post-exit, is key to keeping your options open, says digital sales executive Shannon Leiro. She met regularly with a mentor when she stepped away for two years to start a family. She also monitored industry trends closely and, when she noticed a major shift, determined she couldn’t afford to prolong her hiatus.
“I was like, ‘Oh, my gosh, I better get back in before I become obsolete,’ ” Leiro says.
Chika Uwazie, former chief executive of a technical recruiting startup, thinks she has more runway. She has spent the past few years creating content, including a podcast, newsletter and forthcoming book, “Timeline Grief,” about failing to hit milestones on schedule.
Still, whenever someone reaches out to gauge her interest in returning to the recruiting industry, she makes time for a conversation, even if she doesn’t expect it to go anywhere.
“I always take calls,” Uwazie says. “I think if you’re building your contacts, you can go back. I think the issue is when you are so focused on the new thing that you become disconnected.”
One caution: You don’t want to earn a reputation as someone who leads people on or wastes their time.
But if you are honest about your intentions, peeking back through the door you left open can help ensure it doesn’t slam shut forever.
