Mega Funding Rounds Ignite the Age of AI Behemoths

These days there's so much AI industry news I genuinely don't know where to start — but if you just look at the numbers, you get a pretty good sense of where this industry stands right now: OpenAI just closed the largest private tech funding round in history, $122 billion, at an $852 billion valuation. At the very same time, thanks to the AI agent boom, Anthropic — founded by former OpenAI employees who walked out five years ago — keeps setting new performance records, with annualized revenue now past $20 billion, firmly cementing its place as a first-tier AI player. Five years ago, this company didn't even exist!

Doubling Down: Every Big Tech Company Is Scrambling to Fund AI

Start with OpenAI's round. $122 billion — that number alone makes it the largest private tech funding round in history; Taiwan's entire annual government budget is only about $100 billion. SoftBank, Microsoft, and a raft of venture firms all piled in; weekly active users hit 900 million, and annualized revenue is racing toward roughly $13 billion.

OpenAI CEO Sam Altman noted that when ChatGPT first launched, it took five days to hit a million users, which he thought was already a miracle — but last year, there was an hour when it added a million users. OpenAI had barely shut down its video AI system Sora, with online commentators piling on the doom talk, before turning around and closing the largest single funding round in tech history.

What does that feel like? It's like watching a rocket that's already cleared the atmosphere, and someone's still running up to refuel it — because apparently it's not flying fast enough yet.

But there are some genuinely interesting details buried in this round. Amazon's investment is contingent on an eventual IPO, which is also part of why OpenAI shut down its video platform Sora to concentrate its firepower. Keep in mind, Sora was a pioneering, landmark achievement in generative video AI — much like ChatGPT, it essentially defined what AI in its category looks like. Now OpenAI has cut it loose anyway. The long-term consequence is that Google is left as the near-total monopolist in video AI going forward — but OpenAI clearly can't afford to worry about that right now.

On top of that, Amazon is hedging both sides — it's backing Anthropic and OpenAI simultaneously. If Google weren't sitting on so much cash it doesn't need outside money, presumably Amazon would be trying to get into that round too.

SoftBank actually borrowed $30 billion just to invest in OpenAI — borrowed money, to invest in someone else's company. You read that right. Amazon is issuing corporate bonds too. Practically every major tech company is either raising or borrowing money to fund AI, racing to build data centers. Anthropic closed a funding round back in February, raising $30 billion at a valuation as high as $380 billion — and that's before the AI agent boom really took off. That number is probably a good deal higher by now.

Then there's Oracle, which might be the story that best captures where this industry is really headed. Oracle announced layoffs that could total more than 20,000 people — the reason being it needs to fund its enormous AI infrastructure investment. It's a participant in OpenAI's Stargate project, and its infrastructure commitment there alone exceeds $150 billion, again more than Taiwan's entire annual government budget. So it's laying people off with one hand and pouring money into building more data centers with the other. The money hasn't shrunk — it's just moved from headcount to compute.

Oracle's playbook here is basically identical to what Meta and Amazon have been doing — layoffs and pay raises happening simultaneously sounds contradictory, but the logic is perfectly clear: Oracle is taking the money it used to pay ordinary engineers and redirecting it toward Jensen Huang's Nvidia GPU orders, and toward the small handful of top-tier people who are genuinely scarce in the AI era.

I think this is the clearest signal yet that the AI industry has entered its consolidation phase. In the first half of the race, everyone competed on whose model ran fastest, who had more researchers, who could publish the prettiest papers. Now the competition has shifted to who can build the most compute, who can lock in the most users on their platform, who can simply outlast everyone else. It's moved from a technology race to platform monopoly plus geopolitics — a path the internet industry already walked once, between 2005 and 2015, when Google and Facebook cemented the industry structure that followed (with X playing something like Anthropic's current role). I think the AI industry is walking the exact same path now, just at more than three times the speed.

The Battlefield Shifts: From Technology Race to Platform Monopoly

What does this mean for investing? Short-term noise — things like the standoff between Anthropic and the Pentagon — probably isn't as serious as it looks on the surface, because neither the consumer market nor the enterprise market has actually been meaningfully affected.

Longer term, the gap between the handful of companies capable of sustaining this level of compute investment — OpenAI, Anthropic, Google, Meta — and everyone else is only going to widen. In terms of portfolio positioning, I think it's worth starting to pay attention to who's still standing once this consolidation phase plays out, because the AI world's version of the "search engine wars" is now underway — and that kind of battle usually only produces one or two winners, though because enterprise buyers have more varied requirements, there could end up being one or two extra winners this time.

One more thing: Taiwan's tech industry boom still has a long runway ahead, because when these mega-players raise capital, they're going to spend it buying compute-related equipment — and Taiwan's tech sector is an absolute beneficiary of that. Taiwanese investors should probably be celebrating.