Google on Thursday rolled out a revamped version of its Gemini Deep Research agent, rebuilding the tool on top of Gemini 3 Pro, the company’s latest flagship foundation model.

The new system is designed not only to generate long-form research reports, but also to function as an embeddable research engine that other developers can plug directly into their own applications.

The timing was not lost on anyone in the industry. What once looked like a broad field of AI competitors is becoming narrowed into a head-to-head sprint between two giants that now announce upgrades within the same news cycle, each determined to stay a step ahead in the escalating race for model dominance.

Google’s upgraded Deep Research system is no longer just a tool for generating polished research reports. It has broadened into a programmable research agent suited for large-scale information synthesis, built to survive massive context loads and long reasoning cycles that would normally push an LLM into error territory.

The company is positioning it as a backbone for due diligence review, drug toxicity evaluation, and other high-precision analytical tasks. The major engineering shift was the introduction of Google’s new Interactions API — a feature aimed at letting developers embed Deep Research-grade capabilities into their own products and control the agent layer more directly.

Google also disclosed plans to integrate Deep Research into a series of its core services, including Search, Finance, the Gemini app, and NotebookLM. It is edging toward an internet where people no longer hunt for information themselves; they dispatch an agent trained to do it for them. Google leaned heavily on the claim that Gemini 3 Pro is its “most factual” model so far, tuned to lower hallucination rates in long-chain reasoning tasks where a single incorrect step can ruin the entire output.

To back its performance claims, Google introduced a new benchmark called DeepSearchQA, intended to test how well an agent handles complex, multi-step information missions. It was released as open source. The company also tested the system on Humanity’s Last Exam, a notoriously difficult general-knowledge benchmark filled with obscure tasks, and on BrowserComp, which evaluates an AI agent’s ability to carry out browser-based operations.

In addition to the widely-praised Gemini 3, Deep Research topped the leaderboard on Google’s own benchmark and on Humanity’s. But OpenAI’s ChatGPT 5 Pro landed close behind in both tests and overtook Google on BrowserComp. Even those results barely had time to settle. Hours later, OpenAI announced GPT-5.2, which the company said outperforms rivals — including Google — across a wide suite of standard benchmarks.

This back-to-back rollout marked another escalation in a rivalry that has grown unusually public. The industry now watches releases from both companies like heavyweight rounds. Each upgrade arrives with the sense of a reply to the other.

Google introduced a package that pushes its agentic vision further, giving developers a programmable research companion and moving the company closer to agent-based search. OpenAI countered with a model built to strengthen its leadership in reasoning, speed, and multimodal performance.

The speed of these releases also signals how aggressively both companies are trying to define the next phase of AI development. The focus is shifting from raw model power to advanced agents that can plan, browse, run tools, complete tasks over long stretches, and handle increasingly complex workloads with minimal supervision. Whoever builds the most reliable and scalable agent layer stands to redefine how people use software, how information moves across the web, and how organizations make decisions.

For now, Google seems to be pushing deeper into agentic research systems integrated across its own ecosystem, while OpenAI continues to refine its core model lineup and deploy it across consumer and enterprise channels. Each announcement seems to trigger the next. Each benchmark update sets off another. And as these systems continue to evolve in public view, the industry’s center of gravity keeps shifting toward a true duel for AI supremacy.

The Dangote Petroleum Refinery has cut its ex-gantry petrol price to N699 per liter, a reduction of N129 from the previous N828, and a move that is already reshaping competition in a market long dominated by import-driven pricing.

Sources inside the refinery confirmed the adjustment, and the Independent Petroleum Marketers Association of Nigeria (IPMAN) later validated it.

“The refinery reduced the price to N699 per liter and told marketers that we can start buying the product for the festive season,” IPMAN’s national president, Abubakar Garima, told TheCable.

The price cut has added to the growing shift in Nigeria’s petroleum market. For decades, petrol prices in Nigeria were heavily anchored to the cost of imported products. Dangote’s arrival has tilted that equation. By lowering its ex-depot price to N699, the refinery is setting a benchmark that importers cannot match under current global and domestic conditions.

Aliko Dangote, who has repeatedly expressed the refinery’s goal to dominate the local market, is competing directly with importers.

“Prices are going down. The reason why prices have to go down is because we have to also compete with imports,” he said, adding that both diesel and petrol “will continue to be sold in the market at a very reasonable price.”

The cut is already putting pressure on the Nigerian National Petroleum Company Limited (NNPC), which reduced pump prices across its Lagos and Abuja retail stations in late November to N900 and N940 per liter. Industry sources say that although the NNPC and independent marketers will be forced to make further adjustments once Dangote’s new price filters through the market, they will not be able to compete.

Nigeria’s long-standing dependence on imported petrol is mired in exorbitant import costs. Even when global prices eased, Nigeria faced elevated landing costs because of the weaker naira, freight charges, insurance, and port fees. During late 2024, the landing cost oscillated between N829 and N976 per liter. In November 2024, figures widely circulated by market operators, including the Major Energy Marketers Association of Nigeria (MEMAN), put the landing cost in the N970s. By October 2025, updated industry data showed a drop to the low N830s as global prices cooled and shipping routes stabilized.

Those figures explain why imported fuel cannot compete with Dangote’s new pricing. At N699, the refinery is offering a product more than N130 below the best-case import landing cost and nearly N280 below the high end seen in late 2024. This leaves NNPC and independent importers in no position to bring in foreign petrol and sell at affordable rates without incurring heavy losses.

Against this backdrop, Nigeria’s petrol import is expected to drop further. Nigeria spent N1.28 trillion on petrol imports in the third quarter of 2025, down from N2.3 trillion in Q2, but it remains a major burden on foreign exchange reserves. The drop in Q3 was tied to lower global price averages and occasional improvement in refined product availability. Nigeria’s import bill remains large because domestic output has only recently begun to scale.

In 2020, petrol imports cost N2.01 trillion. By 2021, that figure rose sharply to N4.56 trillion. In 2022, it hit N7.71 trillion. In 2023, it eased slightly to N7.51 trillion. Then came the surge in 2024: N15.42 trillion, driven by a more than 40 percent slide in the naira and higher global shipping and insurance fees.

These costs created long-standing volatility in Nigeria’s retail petrol market, forcing pump prices to chase global dynamics. With Dangote offering large volumes priced locally, that structure is beginning to shift. The refinery’s latest cut puts pressure on all fuel retailers — including NNPC — to align with domestic, rather than imported, pricing.

That shift is happening as Dangote pursues an ambitious expansion plan. The refinery, currently rated at 650,000 barrels per day, is already producing diesel, aviation fuel, and petrol. Work is ongoing to raise total output to 1.4 million barrels per day within three years, a scale that would make it the world’s largest single-site refinery.

The Federal Government has welcomed this expansion, calling it a “game-changer” with the potential to lift West Africa’s entire energy landscape. Beyond stabilizing domestic supply, the government sees Nigeria becoming a regional export hub, with refined products heading to neighboring countries and beyond.

There is also reprieve for one of the refinery’s biggest challenges. OPEC’s latest data shows Nigeria pumped 1.436 million barrels per day in November 2025, up from 1.401 million barrels per day in October. That trend gives the country more room to allocate crude for domestic refining.

Meanwhile, Nigeria’s national petrol consumption has slowed. Federal data shows average daily usage fell to 52.9 million liters in November 2025. Analysts say this is linked to earlier high pump prices, changing transport habits, and the gradual entry of Dangote’s products, which are expected to stabilize supply and lower retail prices in 2026.

With the N699 price now in play, traders say the market has entered a new phase. The days when imported fuel determined domestic pricing are fading. As Dangote ramps up production in 2026, the downstream sector is shifting toward a new price anchor — one set inside Nigeria rather than offshore. However, energy analysts warn that this will be impacted by international oil prices.

Aliko Dangote has unveiled one of the boldest ideas in his business playbook career, and it has little to do with building factories. Instead, it is about re-engineering Nigeria’s capital market through a play of currency. Dangote has announced plans to list 10 percent of his $20 billion refinery on the Nigerian Exchange in 2026, a move that could fundamentally reshape the market and attract global investors seeking hard-currency returns.

He revealed this at the unveiling of the Dangote Vision 2030 roadmap in Lagos, outlining a long-term strategy centered on massive revenue growth, aggressive expansion in petrochemicals, and most notably a shift toward dollar-denominated dividends. The message was unmistakable: Dangote is positioning his companies for far greater international relevance. He confirmed that discussions are already advanced with the Nigerian Exchange and the Securities and Exchange Commission to finalize the structure of the offering, alongside engagements with the Ministry of Finance to secure approvals for the proposed dividend framework. As he put it succinctly: “You buy in naira, but you get dividends in dollars.”

If this plan succeeds, it will trigger a tectonic shift in Nigeria’s capital market. It will set a new benchmark that leading Nigerian banks, many of which already earn substantial dollar revenues through their African subsidiaries, will be forced to match or risk losing investor interest. In a market where capital seeks protection from currency risk, USD-denominated dividends will become a powerful magnet. If Dangote delivers this model, investors will naturally gravitate toward it, compelling others to follow. This is not just a listing; it is a strategic redesign of how value is captured and distributed in Nigeria’s financial markets.

It may not be bad for Naira provided the equities are paid for in Naira even as the dividends come in USD.

“You Buy In Naira, But You Get Dividends In Dollars”: Dangote Affirms Plan Of Refinery’s Historic NGX Listing In 2026

Aliko Dangote has opened a new chapter for his industrial empire with a sweeping plan to list a 10 percent stake in his $20 billion refinery on the Nigerian Exchange in 2026, a move that could reshape Nigeria’s capital market and pull in global investors hungry for hard-currency returns.

He affirmed the plan at the unveiling of the Dangote Vision 2030 roadmap in Lagos, where he laid out a long-term strategy built around massive revenue growth, aggressive expansion in petrochemicals, and a bold shift toward dollar-denominated dividends for shareholders. His message throughout the evening carried the tone of a man positioning his companies for far larger international visibility.

During the presentation, Dangote said his team is already deep in talks with regulators at both the Nigerian Exchange and the Securities and Exchange Commission to finalize the structure of next year’s public offering. He noted that the initiative also includes ongoing engagement with the Ministry of Finance to secure the necessary approvals for the proposed payout framework.

“You buy in naira, but you get dividends in dollars,” he said, adding that the final go-ahead for this framework would come from the Finance Minister.

The dollar-dividend plan is anchored on the refinery’s growing export capacity. Dangote explained that shareholders across Dangote Cement, the refinery, and the fertilizer business would benefit from returns funded by roughly $6.4 billion in annual export earnings. The bulk of that money is expected to come from sales of polypropylene, fertilizer, and other petrochemical products, which will supply the consistent foreign-exchange inflows needed for payouts in US currency.

Analysts say this could mark a major turning point for the Nigerian market, providing a rare instrument that offers hard-currency income at a time when the local currency has struggled.

The refinery listing also aligns with a broader expansion drive that will define the next phase of the conglomerate’s trajectory. Dangote is targeting revenue of about $100 billion by 2030, a leap from the current $18 billion. The ambition is matched by an equally aggressive pursuit of scale, including a plan to push group-wide market capitalization beyond $200 billion before the end of the decade. Within that horizon, Dangote hopes to place his group among the world’s top 100 companies by value.

For the refinery itself, the numbers continue to climb. The 650,000-barrel-per-day plant is already turning out diesel, aviation fuel, and petrol, and Dangote disclosed that capacity will rise to 1.4 million barrels per day within three years. That would make the project one of the largest single-site refineries anywhere in the world and strengthen its role in Nigeria’s effort to produce its own fuels on a large scale.

He called the planned NGX listing the refinery’s first real step toward opening its ownership to the public, and added that secondary listings abroad could follow later, even though the Nigerian market remains the starting point.

“We want the Dangote Refinery to be the golden stock of the Exchange,” he said.

The refinery’s offering is expected to draw vigorous attention from institutional investors. With capacity climbing, export receipts rising, and a potential dollar-dividend structure on the table, analysts say the transaction could become one of the defining market events of the decade, setting a new standard for how large Nigerian companies court global capital.

However, the decision to list a 10 percent stake has stirred debate among analysts who say the offer is too small for a project of such scale, and signals his intent to retain tight control of what is now the most powerful industrial asset in the country.

A refinery of this magnitude would normally be expected to offer a wider public float that encourages broader price discovery, deeper liquidity, and greater institutional participation. Instead, Dangote’s decision to cap the float at 10 percent is being read as a strategic move to maintain overwhelming command of the asset. Given the refinery’s revenue potential, control of the boardroom translates into control of pricing, export strategy, and supply decisions that will shape Nigeria’s fuels market for decades.

So far, the Nigerian National Petroleum Company Limited’s existing stake in the refinery appears to be the largest. NNPC holds 7.2 percent of the project and has openly stated its intention to raise that stake to 20 percent once regulatory conditions allow. If that purchase proceeds, NNPC would become the largest minority shareholder, creating a new equation in which control of the refinery is split between Dangote’s majority position and the government-backed oil company’s strategic influence.

Looking ahead, the dollar-dividend model remains one of the listing’s main attractions. In a currency environment where volatility has made planning difficult for both corporates and individuals, the prospect of regular dollar income has already begun to generate interest among foreign and domestic funds.

Oracle shares suffered a dramatic decline, plummeting by 13% on Thursday following the issuance of a significantly weaker-than-expected earnings forecast.

This sharp drop, which threatened to erase over $90 billion from Oracle’s market capitalization, triggered a broader selloff across the technology sector. The plunge points to a growing skepticism among investors regarding the immense capital cost and delayed profitability associated with the company’s aggressive, debt-fueled expansion into Artificial Intelligence (AI) infrastructure.

The disappointing financial predictions from the crucial OpenAI cloud-computing partner revealed the uneven and often protracted returns from the nascent technology. While many corporate leaders believe AI is the future, Oracle’s results suggest that the massive upfront investments have so far yielded only limited, visible productivity gains or corresponding revenue spikes.

The Financial Strain of the AI Build-Out

Oracle’s financial health is currently being defined by its commitment to the AI race. The company, long a smaller cloud player, dramatically escalated its profile by securing a massive $300 billion cloud computing deal with OpenAI, beginning in 2027. This high-stakes partnership has tightly linked Oracle’s fortunes to the AI startup, leading to increased stock volatility fueled by concerns that rivals like Google may be pulling ahead of OpenAI.

The core financial concerns center on an astonishing acceleration in capital expenditure (CapEx) that is currently straining the balance sheet:

Oracle shocked the market by announcing that its expected CapEx for fiscal year 2026 is now projected to be $15 billion higher than its estimates just last September. This increase pushes the company’s total planned spending to an unprecedented level over the planning horizon, primarily to fund the construction of colossal data centers necessary to meet the demanding requirements of its AI contracts. The OpenAI agreement alone is expected to utilize 4.5 gigawatts of power capacity over the five-year term of the deal.

The immediate cost of this expansion is visible in the company’s cash flow. Oracle reported burning approximately $10 billion in cash during the first half of its fiscal year due to these AI investments, resulting in a negative free cash flow of around -$13.18 billion in the most recent quarter. To finance this build-out, the company has increasingly relied on debt, holding approximately $100 billion in total debt.

Further dampening sentiment, Oracle missed Wall Street consensus estimates for its total revenue, which came in at $16.06 billion versus expectations of around $16.21 billion. It also issued a third-quarter revenue growth forecast that fell below Wall Street estimates, signaling margin pressure as the investment cycle ramps up.

Credit Markets Signal Caution

The sheer scale and speed of Oracle’s debt-fueled CapEx have injected anxiety into the credit markets. This behavior—relying on significant bond issuance rather than strong operating cash flows—marks a departure for Big Tech companies, though rivals like Meta and Amazon.com have also recently issued over $30 billion and $15 billion in bonds, respectively, to finance their AI infrastructure arms race.

Concerns about Oracle’s debt were immediately reflected in derivatives markets. The cost to insure against Oracle’s default, measured by credit-default swaps (CDS), surged by nearly 12 basis points on Thursday, reaching a minimum of five-year highs. Investors are actively dumping Oracle bonds while piling into these CDS contracts to hedge against potential default risk.

The market value hit to Oracle directly impacted its leadership. Larry Ellison, whose net worth of $276 billion (according to Forbes) is largely derived from his roughly 40% stake in the company, saw his net worth potentially decline by more than $30 billion in the wake of the stock plunge.

Broader Tech Sector Selloff

Oracle’s struggles triggered a sympathetic selloff across the entire AI-focused segment, heightening fears that the ongoing AI market frenzy has become an investment bubble reminiscent of the 1990s dot-com boom. This fear is exacerbated by the circular deals and massive, undetailed future spending commitments, such as those from OpenAI (valued at $500 billion but still unprofitable), which is expected to spend over $1 trillion on AI infrastructure by 2030.

The tech-heavy Nasdaq fell to a one-week low as other AI-related stocks and chipmakers followed Oracle’s trajectory. Key components of the AI supply chain—including Nvidia, Advanced Micro Devices, Micron, Broadcom, and Arm Holdings—all recorded declines ranging from 3.1% to 4.2%.

Despite the immediate negative market reaction, at least 13 brokerages cut their price targets, but some analysts defended the company’s strategy. BofA Global Research analysts argued that the decline represents an investment cycle payment, stating, “The current weakness is more capex investment cycles needed to support demand, with the company paying the price for the abnormal speed in which investment is required to meet current AI demand trends.”

However, the market is demanding greater clarity. Oracle now trades at a forward price-to-earnings (P/E) ratio of 29.56, still slightly above that of rivals like Microsoft at 27.24 and Amazon at 29.06, leaving its current valuation under intense scrutiny until its massive CapEx translates into sustained, high-margin cloud revenue.