The artificial-intelligence trade has spent two years minting fortunes in public equities. In late November 2026, the money is increasingly flowing through a different pipe: the exchange-traded fund. A wave of new launches, product pivots and high-profile portfolio moves is transforming AI from a stock-picking exercise into a packaged asset class — even as Wall Street begins to price the possibility that the buildout stumbles.

The busy calendar is bookended by two events this week. In New York on November 19, Bloomberg Intelligence's Emily Pierce convened a panel at the Bloomberg ETFs in Depth conference featuring Ryan Issakainen, senior vice president and ETF strategist at First Trust; Matt Orton, chief market strategist and portfolio manager at Raymond James Investment Management; and Hepsen Uzcan, chief executive of Deutsche Asset Management Americas. Across the Pacific, Australian issuer Betashares added an artificial-intelligence ETF to a technology range it says now manages roughly A$14.1 billion, according to coverage from SMS Magazine, AdviserVoice and Money Management.

The product pipeline is filling fast

The launches are not confined to one market. EMXETF rolled out a China AI ETF designed to give investors targeted exposure to companies building the country's domestic AI ecosystem — a hedge against the idea that the theme is purely a Silicon Valley story. In the United States, Global X's BKCH — long marketed as a blockchain strategy — is gradually repositioning toward artificial intelligence, according to Seeking Alpha. And WTAI continues to pitch itself as well-diversified exposure to large businesses across the entire AI value chain rather than a bet on a handful of chipmakers.

The pace reflects a broader land grab. Morningstar has documented what it calls an AI fever igniting a boom in thematic ETFs, as fund managers hunt for their next big hit. ETF Stream, meanwhile, has spotlighted a narrower question animating product design: how to isolate the companies that actually benefit from AI capital expenditure — power, cooling, networking, optics — rather than those that merely mention AI in an investor deck.

The Anthropic question

Perhaps the most revealing development is Cathie Wood's decision to add Anthropic exposure across six Ark ETFs, a move first reported via MSN. It highlights a structural limitation of the ETF wrapper: the most important AI companies are private. Public funds cannot simply buy OpenAI or Anthropic, so issuers are engineering indirect routes — through investors, suppliers and partners that hold stakes or stand to benefit from the models.

That workaround has a name in the trade: the photonics blind spot. As MSN's coverage frames it, Anthropic's rise has exposed how much of the AI supply chain depends on optical components and interconnect technology that most thematic ETFs barely own. Investors buying "AI exposure" may be getting concentrated semiconductor beta rather than the diversified value-chain exposure the marketing implies.

Has recent volatility changed the long-term investment case for AI — and where are the opportunities emerging beyond the biggest technology names? That was the central question put to the Bloomberg ETFs in Depth panel.

Concentration risk moves to center stage

The Bloomberg panel's agenda — volatility, concentration risk, the infrastructure buildout and how investors use ETFs to diversify — maps closely to the anxieties now driving flows. The paradox of thematic ETFs is that success breeds concentration: as a theme rallies, the largest names grow to dominate the index, and the fund becomes a leveraged bet on a handful of stocks. A small number of megacap technology companies still drive the lion's share of AI-linked returns, and any vehicle that caps their weight risks underperforming the very trend it was built to capture.

Issuers are responding with breadth. WTAI's pitch rests on spanning the value chain; Betashares' new fund is being positioned within an existing technology suite rather than as a standalone moonshot; and the China-focused EMXETF offers geographic diversification at a moment when Nvidia's China business faces renewed trouble, a dynamic flagged in MSN's reporting on Ark's Anthropic bet.

The credit market is watching — and wagering

The most significant development, however, may be happening away from equities. CNBC reported that large investors are making big bets against AI-related debt — from CoreWeave to Oracle and the so-called Magnificent 7 — through credit default swaps, and that those positions could soon reach the ETF market via dedicated credit vehicles. That would give ordinary investors a way to short the AI buildout for the first time at scale.

The signal is worth taking seriously. Much of the AI infrastructure boom has been financed with debt: data centers, power contracts and GPU fleets underwritten on assumptions of sustained demand. Equity investors have largely ignored that leverage. Credit markets, which get paid a fixed coupon and lose first in a downturn, tend not to.

Nvidia's quarter and the capex trade

Nvidia, meanwhile, topped third-quarter earnings expectations and raised its outlook — albeit amid the fallout surrounding its chief executive, a storyline that has added governance noise to an otherwise strong print, according to Yahoo Finance's coverage of ETFs positioned to benefit. The results underscore the durable point: whatever the volatility, hyperscaler capital expenditure continues to flow, and the beneficiaries extend well beyond one ticker.

That is where the ETF industry sees its next chapter. FT Adviser's argument — that ETFs are a flexible route to active investment — is being tested in real time, as passive wrappers carry increasingly active theses. For investors, the lesson of this cycle may be simple: the AI ETF boom has made it easier than ever to own the theme, and harder than ever to know exactly what you own.