OpenAI’s $6.6 billion funding round could be a Pyrrhic victory for the company whose technology has become synonymous with generative artificial intelligence — one that could ultimately serve as a warning for the burgeoning GenAI industry at large.

In early October, the Microsoft-backed business closed a long-anticipated funding round which left it flush with the cash to train up its latest GenAI models and kick the tires on new products, with participation from investors like Nvidia, Thrive Capital and Softbank bringing its valuation up to $157 billion, according to the company. Even as one of the largest funding rounds by a private company closed, however, investors and industry experts were already discussing whether the $6.6 billion would be enough to cement OpenAI’s future long-term.

While the ChatGPT creator reported a spike in its revenue growth over the past year, it also expects to lose approximately $5 billion by the end of the year — with losses related to expenses including the costs of running its products, according to financial documents seen by The New York Times.

The business has also experienced turmoil in its leadership team during the funding round. While CFO Sarah Friar was a top driver for the funding round, top OpenAI executives including its chief technology officer and two senior researchers departed the company in September, days before its funding round closed. Additionally, OpenAI faces growing competition for its ChatGPT products from businesses with deep pockets that rival its main backer Microsoft — which also participated in the most recent round — like Google parent Alphabet and Mark Zuckerberg’s Meta.

The funding round exposed several concerns with OpenAI’s business, with questions arising surrounding funding, potential challenges to revenue and profitability, return on investment challenges and lingering issues regarding data ownership and privacy, experts told CFO Dive. As the best-known, and likely the best-funded, GenAI company, the hurdles OpenAI is facing could mean more than just stormy seas for the company itself: they could indicate dark clouds for the overall GenAI space.

For many companies, the task of divining which way the wind is blowing will fall to their CFOs, tasked with deciding where to spend their companies’ funds — and for many finance chiefs, GenAI’s shine has been dulling for months.

The funding wheel keeps turning  

To be sure, the spotlight on GenAI has yet to dim significantly. Apart from the $6.6 billion OpenAI just pocketed, funding for AI startups reached $19 billion in the third quarter, representing 28% of all venture dollars, according to Crunchbase data. Meanwhile, the number of U.S. companies willing to invest $10 million or more in GenAI is set to double in 2025, CFO Dive recently reported, citing data from Ernst & Young.

But while the funding environment has remained strong, investors — and the CFOs looking to potentially utilize the technology — are starting to ask deeper questions of the companies in their sight lines before drawing out their checkbooks.  Many companies have popped out of the woodwork “positioning what they had been doing, using the more recent nomenclature around AI, machine learning, etc,” said Mark Schwartz, IPO and SPAC advisory leader for Big Four firm EY. “But how much are they doing that's different, relative to …some of the algorithmic-type stuff that they have been doing in the past?”

Understanding that separation is critical for investors as they mull which AI-driven companies are deserving of their capital. “Investors are very keen to understand the technology, to understand if it's truly new, truly novel, and truly potentially revolutionary, as opposed to more of the same,” Schwartz said in an interview.

The need to isolate truly revolutionary companies is beginning to be reflected in venture funding: The $19 billion funneled into AI startups represents a drop from the $23.4 billion seen in the prior quarter, according to Crunchbase, while notably, deal flow for Q3 declined 22% to 947 from the 1,211 rounds seen in Q2.

That “likely means more big deals for more proven AI startups (or at least those that can convince investors of their AI chops) and fewer early seed and Series A deals for young startups,” Crunchbase reported.

“I think underneath the giants there, there's a lot of due diligence that's happening,” Schwartz said of how investors are thinking about GenAI. Some companies are being marked as those that have a differentiator, which are still seeing robust attention in the private markets, but, “there are other companies that, as investors are looking under the hood, they are struggling,” he said.

Crucially for those startups, the world’s largest GenAI player may be proving to be a lackluster poster child. With the company expected to log a $5 billion loss this year, its $6.6 billion round may just be one in a long potential line of raises — sopping up venture dollars and creating more wariness for investors approached by other AI-driven companies.

In fact, given its current burn rate, it’s likely OpenAI will need to complete a similar funding round in 2025 — at a higher valuation than its current $157 billion, tech reporter Ed Zitron predicted in an Oct. 2 newsletter.

“I would imagine, if you're a startup trying to throw AI out there to get huge funding rounds, you’re pissed at Sam Altman,” Matt Cotter, CEO of AP automation software provider Pairsoft said. “Because those guys keep cashing checks, and now everybody turns and looks at you and says, ‘Wait a minute. They're already there, and now I can see their financials.’”

ChatGPT, show me the money

Thanks to its new funding, OpenAI now has access to approximately $10 billion in liquidity, comprised of the $6.6 billion round and a $4 billion credit facility the ChatGPT operator inked with participating banks including Goldman Sachs and JPMorgan, the company announced. OpenAI is also anticipating its revenues will continue to spike, estimating it will hit $11.6 billion next year — compared to expected annual sales of $3.7 billion this year, The New York Times reported.

However, that revenue growth goes hand in hand with rising costs, which represents a key challenge as the company targets that $11.6 billion in annual revenue.

GenAI companies such as OpenAI also face expanding costs related to customer capture, computing power and infrastructure costs. Computing power remains one of the largest cost sinks for OpenAI, and it increasingly needs more than even its top backer Microsoft may be able to supply — Microsoft has not supplied the company with enough power quickly enough, leading OpenAI to approach other potential sources like Oracle, according to reports. 

Part of the issue is that the company’s main cost driver is also its main revenue generator. Presently, the company’s biggest generator of revenue are the subscriptions it sells to its ChatGPT products, with about 10 million users shelling out about $20 per month for such access — a price the company expects to hike to $44 over the next half decade, the NYT said.

“I think what somebody's got to figure out how to do, and I don't think anybody's figured it out yet, is how to change the ratio,” Cotter said. “Revenue keeps doubling and tripling and it's super attractive, but so does cost, because we haven't cracked the code on, how do I deploy all this infrastructure. How do I deploy all this computing power? How do I deploy all this talent?”

Overall, OpenAI expects ChatGPT to generate $2.7 billion in revenue this year. Increasing its subscribers for ChatGPT would mean more revenue, but the sheer computing power needed to generate queries for those users will quickly turn new users into cost centers, Zitron wrote in his Oct. 2 newsletter.

Its other main source of revenue is selling its products to enterprises, where it expects to generate $1 billion by the end of this year. However, that enterprise figure is perhaps the biggest black mark against the company’s future, and by extension, against the monetary draw of the technology itself.

“This suggests a fundamental weakness in the revenue model behind GPT, as well as a fundamental weakness in the generative artificial intelligence market writ large,” Zitron wrote. “If OpenAI cannot make more than a billion dollars of revenue off of this, then it’s fair to assume that there is either a lack of interest from developers or a lack of interest from the consumers those developers are serving.”

Rising prices here could also worsen what Zitron has termed as a coming “subprime AI crisis,"  where “thousands of companies have integrated generative AI at prices that are far from stable, and even further from profitable,” he wrote in a Sept. 16 newsletter.

The attention paid to OpenAI has also prompted companies to funnel more of their budgets into the technology, driving up capital expenditures in a way that may not be sustainable for those smaller companies. 

OpenAI “has been a central, key foundational linchpin to the AI thesis with ChatGPT being the "iPhone Moment" for the world with the introduction of generative AI to enterprises and consumers,” Dan Ives, managing director and senior equity research analyst at Wedbush Securities said in an Oct. 3 note.

“Now we are seeing a $1 trillion of AI Capex over the next three years as OpenAI has been the linchpin to AI success and adoption we are seeing at Nvidia, Microsoft, Google, and across the tech world,” he wrote.

Keeping the eye on the ROI prize

The close attention investors, industry leaders and technologists are paying to OpenAI’s revenue model speaks to a growing need for specifics, rather than potential.

CFOs and their C-suite fellows want both a clear use case from the technology, and, more notably, ways to ensure they’re seeing a positive return on investment from their AI spending. Analysts like David Cahn, a partner at Sequoia Capital, have urged caution in the AI gold rush, with those who remain “level-headed” able to build stable companies, CFO Dive previously reported.

For investors, “what we're hearing is at a super high level, there is a lot of, I'll say, unknowns with respect to just how powerful and just how revolutionary this can be,” Schwartz said.

“I think that this AI revolution has the potential to transform anything and everything top to bottom. I think that right now, we are in the discovery phase of trying to figure out how impactful it can be,” he said. However, “the lack of clarity on valuations in the market” currently is indicative that the space is still in a nascent phase, where leaders are still figuring out how to best move forward, he said. 

For finance chiefs, the GenAI conversation has steadily shifted away from integrating the technology less like a silver bullet, and more like a patch —  something that can solve a particular issue and is being developed accordingly. Indeed, “a focus on practicality, a focus on targeted solutions is going to be, for the world that I live in…a strategy that wins the day,” Cotter said.

Remaining level-headed is one thing — squeezing the ROI investors want to see out of GenAI, in the time frame they want to see it, may be quite another, especially as GenAI companies themselves face challenges growing their revenue where they need to be. Developing those targeted solutions takes a lot of time, although “eventually they're going to be stickier,” Cotter said. “Eventually you're going to get a premium price for them. The problem is, they don't happen in a month.”

Even as leaders isolate specific use cases and projects, “I still don't think that argues that somebody's cracked the code of, ‘how do you change those ratios of revenue to cost?’” Cotter said. “If I'm taking in $6.6 billion in funding, ‘give it time’ is probably not the argument that anybody wants to see here.”

Data coal for the GenAI fire    

OpenAI is already facing a number of challenges, but its last hurdle may be the one that sees it stumble the hardest: data. While the company is facing profitability challenges, the lack of a tidy profit isn’t necessarily a death knell for a business, especially in the technology space; the “move fast and break things” approach proved effective for companies like Uber, which rode to success on wave after wave of venture capital before ever posting a quarter in the black.

But for every Uber, there’s a WeWork: while OpenAI could potentially attract more funding, giving it the time it needs to change its cost to revenue ratio to something more balanced, it could face a supply issue. Simply put, to do all of that, the company needs data. 

ChatGPT and other large language models need access to hordes of data to grow, but gaining access to it isn’t as easy as it might sound. For one thing, with the current training needs of LLMs, “the industry’s need for high-quality text data could outstrip supply within two years, potentially slowing AI’s development,” the Wall Street Journal wrote in an an April report. 

Taking aside the issue of sheer supply, GenAI companies also face an issue accessing much of the data that is available now. Many companies don’t keep their information in formats that are easily accessible for AI tools, meaning they need to set aside more time and money to re-format that information. Furthermore, OpenAI has already been slammed with multiple lawsuits surrounding its access of protected information, raising questions of data privacy and ownership as GenAI usage expands.

“Data provenance is going to be the thing that is existential to OpenAI,” Cotter said. “It is the coal that fuels the plant. if it were to get cut off, legally, technically, something like OpenAI is dead…it thrives on data, period. The process for getting it cut off, though, is murky at best.”

Still, the data question is just one more open-ended problem OpenAI needs to solve to lay the fears of investors to rest. Ultimately, the fate of OpenAI could also be divorced from that of the broader GenAI space, with other players rising to steal the “iPhone moment” away from the ChatGPT creator.

But for finance chiefs, what the cracks here could indicate is that a close, detailed audit of GenAI may be past due — and maybe, one that should be conducted the old-fashioned way.