The AI money moved in two directions this week, and the gap tells you something. On one side, private inference compute funding hit a record, as a single platform raised $1.5 billion in a single round. On the other, public AI stocks in Asia cratered hard enough to trip a circuit breaker. Both happened within 48 hours of each other. If you run a company, the split matters more than either number on its own. Investors keep pouring cash into the plumbing that runs AI models. At the same time, they question what the public market has paid for the chips underneath. Here is what changed, and what it means for your team.

Record Inference Compute Funding Lands at a $13 Billion Valuation

Baseten, a platform that runs AI models in production, raised $1.5 billion in a Series F round announced June 22. The deal closed across two tranches, at valuations of $13 billion and $11 billion. Altimeter, Conviction and Spark Capital led the round. (Source: Business Wire). The company says revenue grew about 20 times year over year. It now handles more than one billion inference requests a day, across 87 clusters and 18 clouds.

This is one of the largest infrastructure rounds of 2026. It also signals where the inference compute funding is going. not into building bigger models, but into serving them fast and cheaply once they exist.

Why this inference funding matters for your team

Most companies do not train models. They run them. Every time your support bot answers a ticket, that is an inference call. Your recruiting tool screening a resume is another. And someone pays for the compute behind each one. So when a single vendor processes a billion of those a day, the cost per call keeps falling. For a 50-person startup, that is the difference between an AI feature that pencils out and one that does not.

The practical read: the tools you buy this year will lean on this kind of infrastructure, whether you see it or not. Therefore, ask your vendors how they handle inference at scale, because their answer now affects your bill later.

What to do: If you are evaluating an AI HR tool, ask where inference runs and how pricing scales with usage. Teams already using AI agents for HR will feel cheaper compute first.

Asia’s AI Stocks Crash, Then Bounce

While private money flooded in, public AI stocks went the other way. South Korea’s Kospi fell about 10% on June 23 and tripped a circuit breaker, a 20-minute trading halt. (Source: CNN Business). SK Hynix and Samsung, which together make up roughly half the index, dropped more than 12%. Meanwhile, Japan’s Nikkei slid 3.6% and SoftBank lost 15%. The next session, the Kospi clawed back about 3%.

So what? This is the AI-valuation jitter showing up in real markets. For founders, it is a useful reminder. the chips and the capital behind your AI tools sit on top of sentiment that can swing 10% in a day. It does not change what you ship Monday. But it does affect how cheaply you can raise if your pitch leans hard on an AI story.

AI Is Now the Top Priority on the HR Desk

A 2026 survey of HR executives found AI adoption across HR tasks climbed to 43%, up from 26% two years earlier. (Source: PR Newswire). Recruiting leads the use cases. The barriers, however, are human, not technical: about 19% cite employee fear of job loss and roughly 17% point to data and compliance worries.

So what? If you lead HR, the adoption question is settled. your peers are already in. The harder work is trust. Nearly half of teams have not set up a way to measure whether AI actually makes them more productive. So start there. Pick one workflow, like AI in HR recruitment, and measure before and after.

Japan Commits $2.3 Trillion, With a Third for AI and Chips

Japan’s Prime Minister Sanae Takaichi unveiled a plan to invest more than 370 trillion yen, about $2.3 trillion, through fiscal 2040. (Source: The Edge / Nikkei). Of that, 101.6 trillion yen is earmarked for AI and semiconductors, nearly a third of the total. The country wants to lift domestic chip sales from about 8 trillion yen a year to 40 trillion by 2040.

So what? Sovereign money on this scale changes where AI infrastructure gets built, and who hires for it. For companies hiring across Asia, that means stiffer competition for chip and AI talent in the region. If you plan to hire there, an employer of record in Singapore and nearby markets can get you compliant faster than setting up a local entity.

Quick Hits

• Qualcomm agreed to buy AI software startup Modular for about $3.9 billion in all stock, picking up the Mojo language and the MAX inference engine to build out its data-center stack. (Source: Network World)
• A US executive order signed June 2 sets a voluntary 30-day government pre-release review for frontier AI models, with no mandatory licensing. It is already shaping rollouts, as the latest GPT-5.6 model ships to government-vetted customers first. (Source: The White House)
• An AI Agent Index presented this week at the FAccT 2026 conference found only 9 of 30 deployed agents publish capability benchmarks, and most skip safety evaluations entirely. (Source: arXiv)

The throughline this week: inference compute funding is reshaping the cost of every AI feature you buy, while the public market argues about what it is all worth. If global hiring and payroll are part of how you scale, Asanify handles multi-country payroll and compliance out of the box. That includes AI payroll automation. Worth a look as the underlying costs shift.

Inference Compute Funding: Quick FAQ

What is inference compute, and why is it drawing record funding?

Inference is the act of running a trained AI model to produce an answer, as opposed to training the model in the first place. Inference compute funding is flowing to companies that serve those models fast and cheaply at scale. Most businesses run models rather than build them, after all. One platform alone raised $1.5 billion for this in June 2026.

Does the Asia AI stock selloff mean the funding boom is over?

Not yet. Public AI chip stocks in South Korea fell about 10% in a single day in June 2026. Yet private investment in AI infrastructure kept setting records the same week. The two markets are sending different signals, so treat short-term stock swings as sentiment, not a verdict on demand.

What should HR leaders do as AI adoption reaches 43%?

Pick one workflow, measure it, and expand from there. A 2026 survey found AI use in HR tasks rose to 43%, from 26% two years earlier. Yet nearly half of teams cannot yet measure the productivity gain. Starting with a single use case like recruiting gives you a baseline before you scale.

Not to be considered as tax, legal, financial or HR advice. Regulations change over time so please consult a lawyer, accountant  or Labour Law  expert for specific guidance.