OpenAI's Unreleased AGI Paper Could Complicate Microsoft Negotiations
Jun 27, 2025 3:35 PM The partnership between OpenAI and Microsoft in many ways hinges on the definition of artificial general intelligence, creating a tension that has spilled over into OpenAI research that has not been made public. Sam Altman, chief executive officer of OpenAI Inc., during a Senate Commerce, Science, and Transportation Committee hearing in Washington, DC, US, on Thursday, May 8, 2025. Photograph:A small clause inside OpenAI's contract with Microsoft, once considered a distant hypothetical, has now become a flashpoint in one of the biggest partnerships in tech.
The clause states that if OpenAI's board ever declares it has developed artificial general intelligence (AGI), it would limit Microsoft's contracted access to the startup's future technologies. Microsoft, which has invested more than $13 billion in OpenAI, is now reportedly pushing for the removal of the clause and is considering walking away from the deal entirely, according to the Financial Times.
Late last year, tensions around AGI's suddenly pivotal role in the Microsoft deal spilled into a debate within OpenAI over an internal research paper, according to multiple sources familiar with the matter. Titled 'Five Levels of General AI Capabilities,' the paper outlines a framework for classifying progressive stages of AI technology. By making specific assertions about future AI capabilities, sources claim, the paper could have complicated OpenAI's ability to declare that it had achieved AGI, a potential point of leverage in negotiations.
'We're focused on developing empirical methods to evaluate AGI progress—work that is reproducible, measurable, and useful to the broader field,' OpenAI spokesperson Lindsay McCallum said in a written comment to WIRED. 'The 'Five Levels' was an early attempt at classifying stages and terminology to describe general AI capabilities. This was not a scientific research paper.' Microsoft declined to comment.
In a blog post describing its corporate structure, OpenAI notes that AGI 'is excluded from IP licenses and other commercial terms with Microsoft.' OpenAI defines AGI as "a highly autonomous system that outperforms humans at most economically valuable work.'
The two companies have been renegotiating their agreement as OpenAI prepares a corporate restructuring. While Microsoft wants continued access to OpenAI's models even if the startup declares AGI before the partnership ends in 2030, one person familiar with the partnership discussions tells WIRED that Microsoft doesn't believe OpenAI will reach AGI by that deadline. But another source close to the matter describes the clause as OpenAI's ultimate leverage. Both sources have been granted anonymity to speak freely about private discussions.
According to the Wall Street Journal, OpenAI has even considered whether to invoke the clause based on an AI coding agent. The talks have grown so fraught that OpenAI has reportedly discussed if it should publicly accuse Microsoft of anticompetitive behavior, per the Journal.
A source familiar with the discussions, granted anonymity to speak freely about the negotiations, says OpenAI is fairly close to achieving AGI; Altman has said he expects to see it during Donald Trump's current term.
That same source suggests there are two relevant definitions: First, OpenAI's board can unilaterally decide the company has reached AGI as defined in its charter, which would immediately cut Microsoft off from accessing the technology or revenue derived from AGI; Microsoft would still have rights to everything before that milestone. Second, the contract includes a concept of sufficient AGI, added in 2023, which defines AGI as a system capable of generating a certain level of profit. If OpenAI asserts it has reached that benchmark, Microsoft must approve the determination. The contract also bars Microsoft from pursuing AGI on its own or through third parties using OpenAI's IP.
Bloomberg previously reported on the existence of the 'Five Levels' paper, and that OpenAI was planning to share the scale with its outside investors, though it was considered at the time as a 'work in progress.' OpenAI CEO Sam Altman and chief research officer Mark Chen have spoken about the five levels of AI capabilities in various interviews since. A version of the paper dated September 2024 viewed by WIRED details a five-step scale for measuring how advanced AI systems are, citing other research that claims many of OpenAI's models at that point were at Level 1, defined as 'An AI that can understand and use language fluently and can do a wide range of tasks for users, at least as well as a beginner could and sometimes better.'
It notes that some models at the time were approaching Level 2, which the authors define as 'An AI that can do more advanced tasks at the request of a user, including tasks that might take an hour for a trained expert to do.' The paper deliberately avoids giving a single definition of AGI, arguing the term is too vague and binary, and instead opts for using a spectrum of capabilities to describe increasingly general and capable AI systems.
The paper doesn't predict when OpenAI's systems will reach each of the five levels, but it does predict how each step up in capabilities could change different facets of society, including education, jobs, science, and politics, warning about new risks as AI tools become more powerful and independent. In a podcast with YCombinator president and CEO Garry Tan in November, Altman said that the company's o1 model could be defined as Level 2, and he expects they'll reach Level 3 'faster than people expect.'
Last July, a coauthor of the paper gave a presentation of the research at an internal event where teams highlighted their most important projects for research-wide awareness, according to multiple sources. The research was well received by other staffers, one source added.
Sources also believe that the paper seemed to be in final stages, and the company had hired a copy editor to finalize the work late last year along with generating visuals for a blog announcing the paper. OpenAI's partnership with Microsoft was cited internally as one reason to hold off on publishing the paper, according to multiple sources who spoke to WIRED on the condition of anonymity as they were not permitted to speak to the press. Another source says that discussions with Microsoft were often 'mentioned as a blocker for putting the paper out.'
McCallum said in a comment to WIRED that 'it's not accurate to suggest we held off from sharing these ideas to protect the Microsoft partnership.' Another source familiar with the matter said that the paper wasn't released because it didn't meet technical standards.
'I think mostly the question of what AGI is doesn't matter,' Altman said at a conference in early June 'It is a term that people define differently; the same person often will define it differently.'
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
CNET
12 minutes ago
- CNET
Tariff Impacts Are Real: I Found 13 Companies With Official Price Hikes
Higher prices for a lot of popular products seem inevitable on our current trajectory. James Martin/CNET In a lot of ways and for a lot of products in the US, the biggest impacts of President Donald Trump's aggressive tariff plans are still a ways off in the near future. Still, numerous companies have already hiked prices or said that they will be increased in the near future -- including, most recently, a popular and CNET-approved brand of smart lights. The fact of the matter is that tariffs -- a tax placed on the importing of certain products into a country -- will ultimately cause prices to go up, with Walmart characterizing these eventual price hikes as "inevitable" during its earnings call last month. Given Trump's push to place historically high tariffs on goods from almost every country in the world, you can also expect these price hikes to hit a huge variety of products. This truth has begun to sink in for a lot of Americans, if a recent survey conducted by CNET is anything to go by. According to the results, about 38% of consumers feel pressured to make certain purchases before tariffs cause them to go up in price. About 10% said that they had already made certain purchases out of the hope that they'll avoid a future price hike, and 27% said they had delayed purchases for products that cost more than $500. Overall, these concerns about prices were felt the most around popular tech pieces like smartphones, laptops and home appliances. To help you keep score, I've put together a list of all the companies that have either confirmed or warned of price hikes due to Trump's tariffs. As other companies make such announcements, you can expect new names to be added here. Continue on for all those details, and for more, find out why it's best if you wait on buying a new iPhone. Best Buy Without getting into specifics, Best Buy CEO Corie Barry told the Wall Street Journal late last month that it has already raised prices on certain products as part of its response to the tariffs. e.l.f. Known as an affordable option in the beauty world, e.l.f. announced in late May that it would be implementing a $1 price hike across its product line in response to the tariffs. CEO Tarang Amin claimed that the reaction from customers was positive, on account of the company's transparency. "We're not trying to pull anything over on anyone," Amin told Fortune. "This is exactly what we're facing, and they understand." Macy's Speaking to CNBC in late May, Macy's CEO Tony Spring said that price hikes will be implemented on some products due to tariffs, while also emphasizing that other tactics -- like discontinuing certain products altogether -- will also be a response to rising costs. Mattel Known for brands like Barbie and Hot Wheels, Mattel sounded the alarm over likely price increases during an early May earnings call. While it's unclear how much the toymaker's prices have increased since then, the company told investors that it would be, "where necessary, taking pricing action in its US business," or to put it plainly, raising prices for consumers to mitigate the impact of tariffs. Nikon Camera-maker Nikon will introduce price hikes in response to Trump's tariffs, effective June 23. This move will only target lenses and accessories the company makes and sells, so the cameras themselves are safe for now. "We will be carefully monitoring any tariff developments and may adjust pricing as necessary to reflect the evolving market conditions," a statement from Nikon explained. "We wish to thank our customers for their understanding and know that we are taking every possible step to minimize the impact on our community." Philips Hue Parent company Signify announced that prices for its popular and highly regarded Philips Hue brand of smart lights will see price hikes effective July 1. The company also confirmed that this decision was made "as a direct result of tariffs." "Signify reserves the right to modify prices based on new or additional tariffs becoming effective in the future," the company's official statement explained. Ralph Lauren Sales at the luxury goods retailer Ralph Lauren have apparently remained steady amid recent uncertainty, but the company is still forging ahead with a plan to combat tariff impacts by raising prices more than it had already intended to, according to the Wall Street Journal. Shein and Temu Trump's tariffs have made a notable target of China, hitting the country with a 30% rate only after initially hiking it all the way to 145%. Online retailers like Shein and Temu rely on direct shipments from markets like China in order to offer the rock-bottom prices that made them famous, so it's little surprise now that they've had to raise prices. The Trump administration has furthered the issues faced by these companies by doing away with a rule known as the "de minimus" exception, which used to exclude smaller purchases under $800 from import taxes. With that rule gone, Trump's China tariffs will now apply to both bulk orders of industrial building materials and those shoes you've been looking to buy from Shein. Subaru Subaru has hiked prices across almost its entire line. The increase ranged from $750 to $2,055, depending on the model, with only the EV Solterra avoiding any change. As has become a trend with some companies, Subaru avoided attributing the price hikes to Trump's tariffs, citing only the common refrain of "market conditions." Trump has notably disparaged companies that explicitly lay the blame for price hikes on his policies. "The changes were made to offset increased costs while maintaining a solid value proposition for the customer. Subaru pricing is not based on the country of origin of its products," a Subaru spokesperson said in a statement to Car & Driver. Stanley Black & Decker In an earnings report published April 30, toolmaker Stanley Black & Decker addressed "Price Actions in Response to US Tariffs," stating that it had "implemented an initial price increase in April and notified our customers that further price action is required," and was also looking into ways to shift its supply lines to minimize the impact of tariffs. Volvo The price impact of tariffs at Swedish automaker Volvo are confined, for now, to just one model: the electric EX30. Initially it was set to start at $34,950 in the US -- a competitive price for an EV -- but tariffs targeted at imported cars forced the company to raise the price to $46,195, a 32% bump. Walmart The biggest grocery chain in the US, Walmart is perhaps the most prominent company yet to announce imminent price hikes due to Trump's tariffs. During the company's earnings call in May, CEO Doug McMillan said price hikes would begin by the end of May and impact things like food, electronics and toys. For more, see why buying refurbished tech helps you dodge tariffs and helps the planet.
Yahoo
19 minutes ago
- Yahoo
Rob MacLennan Named Racing Secretary For Saratoga Meet
Rob MacLennan Named Racing Secretary For Saratoga Meet originally appeared on Paulick Report. The New York Racing Association, Inc. (NYRA) today announced that Rob MacLennan will serve as Racing Secretary during the upcoming summer meet at historic Saratoga Race Course. MacLennan joined NYRA in April 2021 as a racing official and was subsequently named assistant racing secretary before taking on his current role as racing secretary in joint capacity with NYRA's senior director of racing Keith Doleshel in October 2024. With Doleshel focused on other key racing department responsibilities, MacLennan will be charged solely with overseeing the writing of condition books and related duties at Saratoga. Advertisement The 2025 summer meet at Saratoga Race Course will open Thursday, July 10 and continue through Monday, Sept. 1, offering 18 Grade 1 races as part of 64 stakes worth more than $20 million in total purses. The 40-day meet is highlighted by the 156th renewal of the G1, $1.25 million DraftKings Travers on Aug. 23 and the G1, $1 million Whitney on Aug. 2. Following the four-day opening week, racing will be conducted five days a week, Wednesdays through Sundays, apart from closing week, when the 2025 summer meet will conclude on Labor Day. For the complete Saratoga stakes schedule, visit This story was originally reported by Paulick Report on Jun 27, 2025, where it first appeared.
Fast Company
21 minutes ago
- Fast Company
Why your brain matters more than ever in the AI age
It can be enthralling to watch artificial intelligence models progress toward a mastery of deep learning. But are we as equally invested in our own abilities to think and learn? The human capacity to think deeply, find meaning, and apply wisdom is what makes us unique. Yet, it is increasingly tempting and easy to rely on the fast, accessible answers that AI provides. In a recent McKinsey study of organizations that use generative AI, only 27% said that employees review all content created by gen AI before it is used. One-third of respondents said that only 20% or less of gen-AI-produced content is checked before use. The antidote in this moment is critical thinking. Critical thinking is sometimes called 'careful thinking,' as it involves questioning, interpretation, and discernment. Critical thinking is not always our default mode, and it's already under siege from frequent AI usage. However, critical thinking skills can be taught. Moreover, according to our latest research, leaders with strong critical thinking skills have better outcomes, such as confidence in their ability to lead and lower burnout. Thinking Slow or Not at All Whether it's a matter of being lazy or economical, humans don't think a lot if we don't have to. This isn't necessarily a bad thing. Researchers estimate that our conscious brains process information at a rate of 10 bits per second. (AI models process data at trillions of bits per second.) So, we conserve our limited mental horsepower for complex tasks rather than 'wasting' it on simple or repetitive tasks. This is why we go into autopilot mode when we drive familiar routes or rely on mental shortcuts to make decisions. (For example, we are prone to judging a person's trustworthiness based on appearance instead of interactions.) Our slow brains have a new, fast friend called AI. That's a good thing, right? It can be. AI can rapidly process vast amounts of information, recognize patterns that lie beyond human reach, and provoke us to consider new angles. AI-based tools will expand our understanding of business performance, team dynamics, market trends, and customer sentiment. But our new friend can also exacerbate our tendency for cognitive laziness. Remember those mental shortcuts we take? In one shortcut, we overtrust answers from automated systems and don't pay attention to contradictory information, even if it's correct. As AI tools become even smarter and slicker—and answers are delivered in highly confident tones—this automation bias can grow. The downside to all of this is the risk of losing one's own capacity for thinking, learning, and reasoning. Guillaume Delacour, global head of people development at ABB, a technology leader in electrification and automation, spoke to us about the importance of critical thinking for leaders in the age of AI. 'One of the big benefits of AI is that it always has an answer—but this is also a major challenge,' he noted. 'It can be too easy to accept the outcomes it generates. Good leaders have always needed critical thinking, but in our AI-enabled workplace, where every question has an instant answer, this skill is even more important.' Are You a Strong Thinker? Critical thinking is the ability to evaluate situations objectively and make informed, well-reasoned decisions. It requires us to consider biases, question assumptions, and incorporate multiple perspectives. With critical thinking, it's like your brain is doing a workout rather than just lounging on the couch. And, like a physical workout, critical thinking requires discipline, self-awareness, and effort. But the payoff is pretty significant. We recently assessed 227 leaders on their level of critical thinking and divided the group into high and low critical thinkers. We assessed how well each group is likely to operate in the new world of AI, as well as their overall experience as a leader. The differences are striking. Leaders Who Don't Think Will Struggle In a world in which answers can come fast and easily, leaders who score low on critical thinking are at greater risk of letting machines do the thinking for them and becoming increasingly less sharp. · Low critical thinkers are 18% more likely to have confirmation bias than high critical thinkers. Confirmation bias is the tendency to look for or favor information that confirms our existing beliefs. · Low critical thinkers are 32% more likely to over-rely on gen AI for answers. · Low critical thinkers are 36% more likely to demonstrate cognitive failures. Cognitive failures are everyday lapses in memory or functioning during situations we normally are on top of, such as forgetting where you put the car keys. Leaders Who Think Will Thrive Strong critical thinkers have a protective shield against the threats of AI. Critical thinking balances the pull toward cognitive laziness and guards against our natural tendencies to accept and rely on what AI tells us. Moreover, these thinkers have a better experience as a leader. · High critical thinkers rate themselves 14% higher than low critical thinkers on their ability to perform well in their roles. · High critical thinkers rate themselves 13% higher than low critical thinkers on their ability to lead others effectively. · High critical thinkers rate themselves 10% higher than low critical thinkers on their ability to lead confidently into the future. Additionally, high critical thinkers report 21% less burnout in their roles and 16% higher job satisfaction. In important ways, thinking can be a secret weapon for leaders, enabling them to be better at and happier in their jobs. Strengthening Your Thinking Muscle The encouraging news for leaders is that critical thinking is not a 'you have it, or you don't' proposition. Each of us can be a critical thinker, but we need to intentionally rewire our relationship to thinking in order to cultivate this vital leadership skill. Here are a few things to try. Think about your thinking. In the course of a day or week, try taking a mental step back to observe how you think. You could ask yourself questions such as: · What is a belief or assumption that I questioned? · Did I change my mind about something important? · Did I avoid any information because it challenged me? · Did I feel uncomfortable in any ambiguous situations? The underlying skill you are practicing here is the ability to observe how you think and to discern what may be influencing your thoughts. Is there a past experience or possible bias that is playing a role? How much does stress or the need for speed factor in? Practice 'why' questions. When looking at a situation, ask yourself why it happened, why it matters, and/or why a particular conclusion was reached. This habit encourages 'second looks' and slows us down to uncover underlying assumptions, potential biases, and hidden logic. This approach not only deepens our understanding but also stretches our ability to evaluate information from multiple perspectives. Make AI your thinking partner. If we are not careful, our predisposition to cognitive laziness will drive us to pick the fast answers that come from AI models versus the deeper mental workout that comes from wrestling with complex ideas or considering underlying assumptions. But that doesn't mean AI can't play a role. When used well, AI tools can be very effective critical thinking coaches, nudging us to consider new angles or refine our arguments. Always make sure you challenge AI by asking questions such as: How did you come up with that result? Why should I believe that what you are suggesting is correct? What questions should I ask to improve my critical thinking? Bigger Comprehension Thinking has always set humans apart—something to be taught, mastered, and celebrated. In 1914, IBM founder Thomas J. Watson declared 'THINK' as the mantra for the struggling machine organization, saying ''I don't think' has cost the world millions of dollars.' We have arrived now at an incredible point when machines can think and learn in ways far surpassing human abilities. There are benefits to this—ways in which AI can make us all smarter. The key is to stay alert and grounded in what is uniquely human: the ability to examine an answer with clarity, to grasp what's around and underneath it, and to connect it to a bigger comprehension of the world around us.
