Europe’s money has already voted on how to own artificial intelligence: the Xtrackers Artificial Intelligence & Big Data UCITS ETF (XAIX, ISIN IE00BGV5VN51) holds roughly €7.5 billion as of August 2026 — the continent’s largest dedicated AI fund by a wide margin — at a 0.35% TER that undercuts nearly every thematic rival. Scale, price and a first-mover position have made it the default answer to Europe’s most-asked thematic question.
Default answers deserve the hardest questions, so this review asks them: what the patent-based selection actually holds (mostly companies you already own), whether an “AI fund” can meaningfully differ from a tech fund at all, and the overlap arithmetic that decides whether XAIX adds exposure or just fees. Verified against DWS and justETF, August 2026.
XAIX key facts
| Fact | Detail |
|---|---|
| ISIN | IE00BGV5VN51 |
| Ticker | XAIX (Xetra · SIX · London) |
| TER | 0.35% — the thematic shelf’s price leader |
| Selection | AI & big-data relevance via patent-activity screening |
| Replication | Physical |
| Domicile | Ireland (UCITS) |
| Use of income | Accumulating (1C) |
| Fund size | ~€7.5bn (Aug 2026) |
The patent screen: how XAIX defines “AI company”
Every AI fund faces the same definitional swamp — marketing departments have made every company an “AI company” — and XAIX’s answer is more objective than most: the index screens for firms whose patent activity concentrates in AI and big-data fields, weighting toward demonstrated research intensity rather than press-release intensity. The intellectual honesty of that approach is real; so is its consequence: patent leadership in AI belongs overwhelmingly to the mega-cap platforms and semiconductor giants — Nvidia, Microsoft, Alphabet, Meta-tier names dominate, joined by the chip complex and enterprise-software leaders. The screen is objective; the output is concentrated familiarity.
Browse the full series in our tools library.
What that means in practice
XAIX behaves like a research-weighted mega-cap tech fund with an AI tilt — high overlap with the Nasdaq-100, heavy intersection with IITU, and meaningful commonality with SMH’s silicon layer. Its genuine differentiators: international patent leaders the US-listing funds skip, and mid-cap AI specialists beneath the platform tier. Those differentiators are real and minority; the majority of the fund is the same ten companies every tech vehicle holds.
Comparing UCITS funds? Take the decision workbook with you.
- The ETF Comparison Workbook — 10 pages: the 7-filter framework for choosing between two good funds
- + 2 bonus PDFs: ETF Starter Kit and Dividend Starter Sheet
- + a free lesson from the €147 Masterclass
The uncomfortable question: can an AI fund exist at all?
The deepest issue with the category deserves plain words. AI is not a sector — it’s a technology diffusing through every sector, which means “AI exposure” collapses into three investable layers: the compute layer (chips — SMH owns it purely), the platform layer (hyperscalers monetizing AI at scale — every world and Nasdaq fund owns it at full weight), and the application layer (software embedding AI — diffuse, early, hard to isolate). XAIX samples all three, which is both its pitch and its dilution: you get the whole stack, already owning most of it elsewhere. The fund’s best claim is convenient concentration — a research-weighted tilt toward the stack in one line at a fair price — not unique access. Buyers who understand that buy it correctly; buyers expecting hidden AI gems are shopping in the wrong aisle.
XAIX vs the AI shelf
- vs L&G AIAI (0.49%): the direct rival — expert-committee selection vs patent screen, purer mid-cap tilt vs mega-cap weight, at 0.14% more. The head-to-head lives in the AIAI review.
- vs the Nasdaq-100 (0.30–0.33%): the awkward benchmark — similar top weights, similar behavior, similar price. XAIX’s case is the patent-screened internationals and specialists; measure whether ~30% of differentiated holdings justifies a dedicated position.
- vs SMH (0.35%): the compute layer alone at identical price — more concentrated, more cyclical, historically the sharper AI-capex expression.
- vs doing nothing: the world-index holder already owns AI’s winners at market weight — the null option every thematic review on this site takes seriously.
Taxes and buying
Irish accumulating UCITS standard (guide, calculator): euro lines on Xetra, wide savings-plan availability at the neobrokers (setup), near-zero dividend yield. At €7.5bn, spreads are thematic-best.
Risks
- Mega-cap concentration: the same handful of names driving every tech vehicle drive this one — single-platform news is fund news.
- AI-capex cycle: the fund’s fate rides datacenter spending; a digestion phase hits the whole stack simultaneously.
- Valuation regime: the theme trades at famous-thesis multiples; 2022 showed the de-rating template.
- Patent-screen lag: research intensity is a trailing indicator — the screen rewards yesterday’s labs, and genuine newcomers enter after their run, the structural cost of objectivity.
- Triple-counting risk — the portfolio-level danger covered next.
The overlap audit: this fund’s real decision
Run the numbers before buying: a €50,000 portfolio holding a world core (~25% effective top-tech weight), plus a Nasdaq or IT-sector fund, plus XAIX can quietly reach 40%+ exposure to the same ten companies through three fee layers. The audit takes ten minutes on justETF’s overlap data and usually returns one of two verdicts: either XAIX replaces another tech vehicle (legitimate — its price and breadth make it a fine flagship tech tilt), or it stacks on them (the accidental concentration this site’s framework exists to prevent). One tech tilt, deliberately chosen, 5–15% ceiling — the rule survives every new theme’s arrival, including this decade’s biggest.
Reader scenarios
“I want AI exposure and own only VWCE — is XAIX the add?” Defensible as the single deliberate tilt: 5–10%, understanding you’re doubling the mega-caps you already own plus adding the specialist tier. The alternative reading: VWCE already gave you AI’s winners at market weight, and the tilt is conviction, not completion.
“XAIX or SMH for the AI decade?” Chips for the sharpest capex expression, XAIX for the whole stack with less cyclicality — the SMH review’s framework applies. Both at once is one bet, twice.
“Why has XAIX grown so much faster than rivals?” Price, size begets size (liquidity attracts flows), and DWS’s distribution — the boring reasons category leaders lead. Flows validate the wrapper, not the thesis.
“What would make XAIX outperform the Nasdaq meaningfully?” The mid-cap AI specialists and internationals outrunning the platforms — historically episodic, never yet sustained. Price the fund as Nasdaq-plus-optionality and the expectations stay honest.
What to watch through 2026–2027
Hyperscaler capex guidance — the stack’s master variable, as across this site’s whole tech coverage. AI monetization data — enterprise software’s AI revenue lines turning material would vindicate the application layer XAIX samples. The fee war — 0.35% leads today; AI fund launches keep coming and price pressure with them. Index reconstitutions — watch which newcomers the patent screen admits; the additions list is the fund’s honest research signal.
Quick-start checklist
- Run the overlap audit first — replace, don’t stack.
- Size the tech tilt once: 5–15% across ALL tech vehicles combined.
- Buy the Xetra euro line; automate monthly — theme entries are timing traps.
- Write the exit condition (capex cycle turn, valuation threshold — yours, but written).
- Rebalance annually; AI’s good years will try to become your whole portfolio.
Verdict
XAIX earned its throne honestly: the most objective selection method in a hype-soaked category, the best price, dominant liquidity. What it cannot do is escape the category’s truth — AI’s investable winners are mostly the giants everyone already owns, and every AI fund is a re-weighting of familiar names plus a specialist tail. As a deliberate, audited, singular tech tilt, XAIX is the shelf’s best default; as a portfolio’s third helping of Nvidia, it’s expensive déjà vu. The satellite framework decides which — before the order does.
What the AI capex cycle means for this fund, mechanically
XAIX’s fortunes route through one spending chain: hyperscalers budget datacenters → chip and hardware orders flow → enterprise software attaches AI features → the patent leaders’ revenues compound. Every link is visible in public numbers — capex guidance each earnings season, semiconductor bookings, software firms’ AI-attach disclosures — which makes this theme unusually auditable for holders willing to read four quarterly reports a year. The cycle risk is equally mechanical: when the buildout digests (every infrastructure boom has), the chain reverses in order — hardware first, platforms cushioned by services, applications last. XAIX’s whole-stack breadth means it feels each phase moderately rather than any phase extremely — the diversification argument inside the theme, and the reason its drawdowns should run shallower than SMH’s while its rallies run flatter.
Three portfolio constructions using XAIX correctly
The flagship tilt: world core + XAIX 8–10% as the sole tech vehicle — the cleanest expression, replacing Nasdaq/sector funds rather than joining them. The layered stack: XAIX + SMH split ~6%/4% for whole-stack plus silicon-sharpened exposure — defensible for conviction holders who accept the overlap consciously. The completionist error: XAIX + CNDX + IITU + SMH — four fee meters, one bet, the audit’s textbook failure case. The middle construction is where most sophisticated readers land; the first is where most should.
Common mistakes with AI investing
- Buying the theme at narrative peaks: AI enthusiasm arrives in waves, and each wave’s top has been a measurably poor entry — automation through the cycle beats every headline-triggered purchase pattern the category has produced.
- Stacking AI vehicles: the audit failure covered above — one fund, summed sizing, or the theme quietly becomes the portfolio.
- Expecting the fund to find hidden winners: patent screens weight proven research leaders; the undiscovered-gem fantasy belongs to venture capital, not UCITS wrappers.
- Confusing AI exposure with AI immunity: the theme’s holdings are exactly the stocks a rate shock or capex digestion hits first — an AI fund raises tech beta, never hedges it.
The theme in one honest paragraph
Artificial intelligence is simultaneously the decade’s most important technology and its most crowded trade — both facts true, neither canceling the other. XAIX resolves the tension the only way an index product can: by owning the measured research leaders broadly, cheaply and mechanically, leaving the investor exactly one job — sizing the exposure against what a world portfolio already carries. Investors who do that job hold the theme sustainably through its inevitable digestion phases; investors who skip it discover during the first real drawdown that their diversified portfolio had quietly become a leveraged bet on ten companies’ capex narratives. The fund is excellent; the audit is mandatory; the order of those two clauses is the entire review.
The high-conviction alternative: the actively managed route into the same stack — Cathie Wood’s European wrapper — is assessed honestly in the ARK AI & Robotics (ARKI) review.
The buildout alternative: investors who prefer AI’s physical invoice — datacenters, power, cooling — over the software stack should read the iShares AI Infrastructure (AINF) review, 0.35% and picks-and-shovels by design.
The vintage label decoded: how big data became the AI shelf this fund flagship-anchors — and the routing table for every layer — is in the big data ETF guide.
Frequently asked questions
Is XAIX actively managed?
No — it tracks a rules-based index using patent-activity screening; the objectivity of the method is its differentiation from committee-picked rivals.
How much of XAIX overlaps the Nasdaq-100?
Substantially at the top — the mega-cap patent leaders are the same names. The differentiated minority lives in internationals and mid-caps; justETF’s overlap tools quantify the current figure.
Does XAIX hold private AI companies like OpenAI or Anthropic?
No — UCITS funds hold listed equities only. The frontier labs are private or subsidiary-held; public-market AI exposure routes through their platform investors and suppliers, which is precisely what the fund owns.
Why is the TER only 0.35%?
Scale and competitive intent — DWS priced for category leadership and got it. Holders simply benefit.
Is there a distributing class?
The 1C flagship accumulates; the sector yields almost nothing anyway. Income lives on the dividend shelf.
XAIX or a robotics fund?
Different layers of the same revolution: XAIX owns the intelligence, RBOT the physical embodiment. The pair overlaps less than expected — but two themes still need two justifications.
How often does XAIX’s index reconstitute?
Scheduled reviews refresh the patent screening periodically — additions lists are worth reading as research signal, since they show where measured AI investment is actually concentrating beyond the headlines.
Does XAIX hedge currency?
No — the mega-cap holdings are dollar-heavy and the euro line carries that exposure unhedged, standard for the thematic shelf and immaterial next to the theme’s equity volatility.
Cited sources: DWS/Xtrackers fund data, justETF data, and index provider information. Verify current fund figures before investing. Educational content, not financial advice.
European UCITS/ETF investor with 11+ years building a global dividend and index portfolio, and founder of Financial Expert Class. I write every guide here from first-hand experience investing across EU brokers and tax regimes — not theory. More about me →