OpenAI is in talks to raise as much as $100 billion in a new funding round that could value the ChatGPT maker at up to $830 billion, according to a Wall Street Journal report on Thursday.

But the sheer scale of the valuation is also sharpening unease across financial markets, as OpenAI’s revenue generation still lags far behind the level of spending implied by such a price tag.

According to the Wall Street Journal, the ChatGPT maker is seeking to complete the funding round by the end of the first calendar quarter next year and is expected to approach sovereign wealth funds, a sign the company is looking for investors with deep pockets and long-term horizons. The Information earlier reported the talks, placing the potential valuation at about $750 billion, still an eye-watering figure for a private company with limited operating history at this scale.

For supporters, the fundraising effort is a powerful vote of confidence. It suggests that large investors believe OpenAI will sit at the center of the AI economy, capturing value across consumer products, enterprise software, developer tools, and foundational infrastructure. OpenAI’s rapid adoption, particularly of ChatGPT, and its growing influence in setting the pace of model development have made it a de facto bellwether for the sector. In that sense, the willingness to commit tens of billions more reflects a belief that AI will reshape productivity, labor, and entire industries, even if profits are still some distance away.

However, the size of the valuation is also raising red flags. At a potential $830 billion, OpenAI would be valued well above many established, profitable technology giants, despite generating a fraction of their revenue and still burning vast amounts of cash. People familiar with the company’s finances say OpenAI is currently generating about $20 billion in annual run-rate revenue, largely from subscriptions, enterprise contracts, and API usage. While that figure is impressive for a relatively young company, it remains small compared with its spending trajectory.

OpenAI’s costs are dominated by compute. Training frontier models and, increasingly, running them at scale for millions of users requires enormous investment in data centers, chips, and energy. The company has signaled plans that could ultimately involve trillions of dollars in infrastructure spending. Inferencing, the cost of serving models to users in real time, is emerging as a particularly heavy burden. Unlike model training, which has often been offset by cloud credits from partners, inferencing appears to be funded largely in cash, meaning operating expenses rise directly with usage.

This imbalance between revenue and expenditure is at the heart of investor concern. Even assuming strong growth, OpenAI’s current revenue base does not yet justify a valuation approaching $1 trillion by traditional metrics. That gap forces investors to price in years, if not decades, of rapid expansion, rising margins, and eventual dominance of multiple AI markets. Any slowdown in adoption, pricing pressure from competitors, or regulatory intervention could challenge those assumptions.

The timing of the fundraising talks also matters. Broader sentiment around AI has cooled as investors also question whether the debt-fueled investment cycle driving the sector can be sustained. Companies such as Amazon, Microsoft, and Oracle have poured tens of billions into AI infrastructure, often ahead of clear near-term returns. At the same time, constraints in the supply of advanced chips, particularly high-bandwidth memory, threaten to push costs higher and slow deployment, adding another layer of risk.

Competition is also intensifying. Rivals, including Anthropic and Google, are accelerating model releases and expanding their ecosystems, forcing OpenAI to spend aggressively to maintain its lead. That pressure has narrowed the margin for error: OpenAI must continue innovating while simultaneously proving it can convert scale into durable, high-margin revenue.

OpenAI has been rumored to be exploring an initial public offering as another way to raise capital, and there has also been talk of courting Amazon for a roughly $10 billion strategic investment tied to access to its in-house AI chips. According to PitchBook, the company already has more than $64 billion in cash and was most recently valued at about $500 billion in a secondary transaction, meaning any new round would represent a sharp step up in expectations.

In the end, the reported fundraising captures the paradox of the AI boom. Investor appetite remains strong, and confidence in the long-term impact of AI is clearly intact. But as valuations soar far ahead of revenues, OpenAI is becoming a focal point for a deeper question confronting markets: how quickly, and how convincingly, can AI’s promise be translated into profits that justify the scale of capital now being committed?

Palo Alto Networks on Friday said it will migrate key internal workloads to Google Cloud under a new multibillion-dollar agreement that significantly broadens and deepens an already extensive strategic partnership.

The deal highlights how cybersecurity and cloud computing are becoming inseparable as companies accelerate the adoption of artificial intelligence.

The agreement expands an existing relationship that has steadily grown over several years, positioning Palo Alto Networks more tightly within Google Cloud’s infrastructure and AI ecosystem at a moment when enterprises are grappling with how to deploy generative AI tools without exposing themselves to new categories of cyber risk.

As part of the deal, Palo Alto Networks is now using Google’s Gemini large language models to power its AI-driven security copilots, tools designed to help security teams detect, investigate, and respond to threats more quickly. The company is also relying on Google Cloud’s Vertex AI platform to build, train, and deploy AI models across its product suite, according to a joint release.

“Every board is asking how to harness AI’s power without exposing the business to new threats,” BJ Jenkins, president of Palo Alto Networks, said in a statement. “This partnership answers that question.”

The expanded collaboration goes beyond internal migration and product development. Palo Alto Networks and Google Cloud said the new phase of the partnership will allow joint customers to protect live AI workloads and sensitive data running on Google Cloud, enforce consistent security policies across cloud and hybrid environments, and simplify complex security architectures that often rely on multiple, disconnected tools.

The focus on securing AI workloads reflects a growing concern among enterprises that generative AI systems, AI agents, and large language models introduce unfamiliar vulnerabilities. These include risks around data leakage, model manipulation, prompt injection, and the exposure of proprietary or regulated information, challenges that traditional cybersecurity tools were not designed to address.

Palo Alto Networks already has more than 75 joint integrations with Google Cloud products and has completed roughly $2 billion in sales through the Google Cloud Marketplace, underlining the commercial importance of the partnership even before the latest expansion. The companies said the new agreement will deepen engineering collaboration, allowing teams on both sides to co-develop security capabilities directly within Google Cloud’s AI and infrastructure stack.

Migrating internal workloads to Google Cloud is also a strategic signal for Palo Alto Networks. The company gains firsthand experience securing large-scale AI systems, an experience it can then translate into products for customers facing similar challenges. The move also tightens Palo Alto Networks’ alignment with one of the world’s largest cloud providers as competition intensifies among security vendors to become the default layer for enterprise AI protection.

For Google Cloud, the partnership strengthens its positioning in enterprise security, a key battleground as it competes with rivals such as Amazon Web Services and Microsoft Azure. Security has increasingly become a deciding factor for large organizations choosing where to run mission-critical and AI-driven workloads, and alliances with established cybersecurity firms are central to that strategy.

“This latest expansion of our partnership will ensure that our joint customers have access to the right solutions to secure their most critical AI infrastructure and develop new AI agents with security built in from the start,” Google Cloud President Matt Renner said.

Market reaction to the announcement was measured. Shares of Palo Alto Networks rose about 1% on Friday, while Google shares were mostly flat, suggesting investors view the deal as strategically meaningful but consistent with broader trends in cloud and cybersecurity convergence.

More broadly, the agreement illustrates how the cybersecurity industry is evolving. Rather than selling standalone tools, leading firms are increasingly embedding their technologies directly into cloud and AI platforms, aiming to become foundational components of customers’ digital infrastructure. As enterprises move faster to deploy AI across operations, the ability to offer integrated, cloud-native security with AI built in from the outset is becoming a central differentiator.

In that context, the Palo Alto Networks–Google Cloud deal is not just an expansion of a partnership, but a reflection of how AI, cloud computing, and cybersecurity are converging into tightly linked ecosystems, with both companies seeking to lock in customers as AI adoption reshapes enterprise technology priorities.

Nigeria’s public finances are coming under increasing strain as debt servicing continues to consume the bulk of government revenue, leaving little room for investment in infrastructure and social services.

New data from the Federal Government shows that in the first seven months of 2025, Abuja spent almost three-quarters of its income paying creditors, a trend that underscores how deeply debt obligations are shaping fiscal outcomes.

The situation is becoming even more troubling when set against the scale of the country’s revenue weakness, with new budget data showing that the government is not only struggling to fund development but is increasingly borrowing just to meet existing debt obligations.

Figures from the 2026–2028 Medium-Term Expenditure Framework and Fiscal Strategy Paper reveal that in the first seven months of 2025, the Federal Government generated N13.67 trillion in revenue, yet spent N9.81 trillion servicing domestic and foreign debt. That meant nearly 72% of all income was swallowed by debt service. When personnel costs of N4.51 trillion are added, spending on wages and debt alone rose to N14.32 trillion, exceeding total revenue for the period.

This imbalance is unfolding against the backdrop of a far deeper revenue crisis. The Federal Government’s 2025 budget was anchored on spending of about N40 trillion, but Finance Minister Wale Edun told members of the House of Representatives Committees on Finance and National Planning earlier this week that the federal government is on course to close 2025 with total revenue of about N10.7 trillion — barely a quarter of the N40.8 trillion projected. That leaves a yawning gap of roughly N30 trillion, highlighting the scale of the fiscal hole confronting the administration.

The implication is twofold. First, the government has been unable to implement large portions of the 2025 budget because expected revenues have failed to materialize. Second, and more worrying, borrowing has increasingly been used not to fund new infrastructure or growth-enhancing projects, but to service existing debt and keep basic government operations running.

Oil revenue remains the weakest link. Between January and July 2025, oil earnings stood at N4.64 trillion, far below the pro rata target of N12.25 trillion. The resulting shortfall of over N7.6 trillion underscores the vulnerability of public finances to crude production challenges, oil theft, and price volatility. Dividends from key entities such as Nigeria Liquefied Natural Gas and development finance institutions also fell sharply short of expectations, further tightening revenue inflows.

While some non-oil taxes showed modest resilience, they were nowhere near sufficient to plug the gap. Company Income Tax slightly exceeded its target, and Value Added Tax performed better than projected, but customs revenue, Federation Account levies, and oil-related inflows all posted steep declines. Overall, aggregate revenue of N13.67 trillion was more than N10 trillion below the pro rata target for the first seven months of the year.

On the expenditure side, debt servicing overshot budget estimates, reaching N9.81 trillion compared with a pro rata target of N8.35 trillion. Foreign debt service, in particular, exceeded projections by nearly 29%, reflecting higher external obligations and exchange rate pressures. This overshoot came even as non-debt recurrent spending and capital expenditure were sharply curtailed.

Capital spending suffered the most. Only N3.60 trillion was spent on capital projects against a pro rata budget of N13.67 trillion, a shortfall of almost 74%. Releases to ministries, departments, and agencies were especially weak, limiting progress on infrastructure, health, education, and other critical sectors. The Budget Office partly attributed this to the extension of the 2024 budget, but the broader issue remains the lack of cash to fund capital votes.

The picture that emerges is of a government boxed in by deficient revenue, rising debt service costs, and shrinking fiscal space. With only a fraction of its N40 trillion budget funded, Abuja has been forced to prioritize debt repayment and salaries over development spending, even as borrowing continues to rise.

In 2024, debt service already consumed 77.5% of the federal government’s revenue. The 2025 figures suggest the situation has not improved. Instead, Nigeria appears caught in a cycle where weak revenues limit budget implementation, borrowing fills the gap, and an increasing share of income is then used to service the resulting debt. Without a significant turnaround in revenue generation, the data points to growing constraints on growth, investment, and the government’s ability to deliver on its policy promises.