Analysis: The AI Market Starts Demanding Proof — From Nvidia's Hugging Face Acquisition to OpenAI's Stuttering Fundraise

For the first time in the years-long AI boom, this week the market asked the question it had long deferred: when does all this spending pay back? The chain of events from 28 October to 1 November 2026 points to a psychological turning point — investors are no longer buying promises, but demanding proof.

Three signals arrived almost simultaneously: the US regulator began sniffing around AI industry consolidation, giant chip and technology stocks tumbled in a mass sell-off, and OpenAI's funding round — which should have been the proof of investor confidence — ran sluggishly far below target.

## DOJ Eyes the Biggest Acquisition in AI

The biggest news came from Reuters on 28 October 2026: the United States Department of Justice (DOJ) is questioning the competitive effects of Nvidia's planned acquisition of the Hugging Face AI model platform, valued at 12.93 billion US dollars.

The regulator's concern is structural, not merely about size. Nvidia already dominates the supply of AI training and inference hardware — the GPUs that run nearly every large model in the world. Hugging Face is the largest open-source model distribution hub, where millions of developers download and share AI models. Merging the two means a single entity controlling both the supply of computing power and the model distribution shelf — in analysts' terms: controlling the picks and shovels plus the only grocery store in town.

Nvidia has promised to keep Hugging Face open and multi-vendor, but trust in that promise is thin. Nvidia declined to comment to Reuters, and the deal is only expected to close in the first half of 2027. The questions the DOJ is reported to have asked include whether Nvidia could use the acquisition to squeeze competitors — for instance, by prioritising certain models or disadvantaging competitors who are also its customers.

Notably, the probe runs alongside a broader investigation into Nvidia's business practices that began in the Joe Biden era and continues under Donald Trump — covering the Groq licensing deal in 2025, chip allocation to customers, and the SchedMD and Run:ai acquisitions.

## The Sell-Off and the Return Test

Almost simultaneously, Nvidia shares fell 2.6 percent on Friday (30/10/2026), dragged down in a broad sell-off across the technology and semiconductor sectors. According to Reuters, negative sentiment was driven by Amazon's aggressive AI spending plans and Intel's warning about large cash outflows, plus a weak outlook.

This is not mere daily volatility. The narrative is clear: the market is starting to question the cost-and-return profile of building AI infrastructure. Lenders are reported to have stopped accepting GPU leases as collateral — a sign that the financial instruments that have underpinned "neocloud" expansion are coming under doubt.

Nvidia itself was not silent. In the same week, the company expanded its cooperation with Meta — covering the Muse AI models and their agents — and pledged to invest 3 billion US dollars in Meta's enlarged share offering of up to 97.5 billion US dollars. Nvidia also expanded its supercluster plans with Oracle to multi-gigawatt scale. These moves show Nvidia becoming ever more intertwined with nearly every node of the AI ecosystem — which only thickens the structural concerns raised by the regulator.

## OpenAI: A Fundraise That Fell Short of Its Promise

Meanwhile, OpenAI — the brightest symbol of the AI boom — experienced what the Wall Street Journal described as a fundraise far below target. From an initial target of 30 billion US dollars, the WSJ reported commitments gathered at below 15 billion US dollars. The Financial Times later reported the round's target was revised to around 20 billion US dollars at a 1.25 trillion US dollar valuation, after several investors cut their commitments amid AI market turbulence.

Ironically, the fundamental numbers are actually strong: OpenAI's run-rate revenue reached 40 billion US dollars in August, up around 70 percent since July. The customers are there, the revenue is there — but investor appetite for financing limitless expansion apparently has limits.

For broader context: Samsung in the same week reported third-quarter operating profit of 107.4 trillion won — up 783 percent year-on-year, the first 100-trillion-won quarter ever recorded by any technology company — with 2027 HBM memory supply already sold out in advance. Demand for AI infrastructure is genuinely there. What the market is questioning is not demand, but who ultimately bears the costs and when the returns arrive.

## The Filling-In Picture

The same week also saw Google selling its TPUs outside its own walls for the first time — Oracle will offer Google TPUs through Oracle Cloud starting in 2027 — and news that OpenAI plans to buy TPUs directly from Google. The AI giants once locked into a single supplier's ecosystem are now diversifying. That can be read as market maturity — or as a sign that Nvidia's dominance is starting to feel uncomfortable even for its largest customers.

Earlier, RingkasNow analysed how Alphabet lost around 500 billion US dollars in market value in a single day after its management admitted slower growth in search advertising tied to AI Mode's lower monetisation. The pattern is the same: the market punishes when AI costs start "eating" existing businesses without commensurate evidence of revenue replacement.

## What It Means for Indonesia

For Indonesia, this wave carries two implications. First, global AI computing costs coming under question could pressure GPU rental prices in the medium term — good news for local AI startups and universities long burdened by cloud costs. Second, the regulator-scrutinised consolidation is relevant to the issue of digital sovereignty: the fewer players controlling the AI value chain, the greater the dependence of countries like Indonesia on business decisions made in Silicon Valley.

One thing is clear: the era of "spend first, calculate later" in the AI industry is coming to an end. The market is now demanding proof — and how Nvidia, OpenAI, and the other giants answer will determine the direction of this industry over the next few years.