The OCF range across the five major AI and technology funds examined here runs from 0.01% to 2.86% — a disparity that has nothing to do with manager skill, portfolio construction or the underly ing thesis on artificial intelligence. It is almost entirely a function of share-class architecture, and it is the first thing any serious investor needs to understand before asking whether these funds are holding the same stocks.
Three of the five funds in this comparison — Pictet – Robotics I USD (LU1279333329), BNP Paribas Funds – Disruptive Technology N Capitalisation (LU0823422141) and BlueBox Funds – BlueBox Global Technology Fund Class R-UK USD (LU2992393228) — report OCFs of 0.01%. BlackRock Global Funds – Next Generation Technology Fund Class A2 SGD Hedged (LU1861220033) sits at 0.02%. These are institutional and specialist share classes, in some cases currency-hedged, with minimum subscriptions that place them well beyond the reach of most private investors. Allianz Global Investors Fund – Allianz Global Artificial Intelligence CT EUR (LU1602091867), at 2.86% OCF, is a retail share class priced accordingly. Comparing these as a fee spectrum would be misleading — they are not competing for the same investor at the same access point.
The convergence question, properly framed
With the fee distortion acknowledged, the more interesting question remains: are these strategies actually doing different things with the capital they manage? The combined AUM is substantial. Pictet’s strategy holds $11.4bn, the Allianz fund €8.1bn, BNP $5.0bn, BlueBox $2.2bn, and the BlackRock SGD-hedged class the equivalent of roughly SGD 2.1bn. Taken together, these represent a significant slice of specialist AI and technology fund capital in European and Asian markets.
The thesis behind each strategy is superficially distinct. Pictet frames its mandate around physical automation and robotics, emphasising the hardware and sensor layer of the AI stack. BlackRock’s Next Generation Technology mandate has historically weighted cybersecurity and data infrastructure alongside hyperscaler exposure. BNP’s Disruptive Technology approach casts a wider net, encompassing fintech and healthcare technology. BlueBox concentrates on high-quality global technology with an emphasis on free-cash-flow compounders. Allianz, the largest retail-facing product in the group, positions itself as a full-spectrum AI play spanning enablers, adopters and beneficiaries.
These distinctions are real. They are also substantially narrower in practice than the mandate descriptions suggest.
Where the overlap is hardest to escape
The structural problem for any AI or technology fund manager in 2025 and 2026 is that the infrastructure layer of the AI stack has consolidated around a small number of companies whose scale is genuinely difficult to route around. Hyperscaler capital expenditure is running at extraordinary levels across the four largest US cloud providers, and the primary beneficiary of that spending — at the silicon layer — remains singular. A robotics fund, a disruptive technology fund and a pure-play AI fund will each arrive at the same name through entirely different analy tical frameworks and still find themselves holding similar weights in the same stock.
This is not irrationality on the part of fund managers. The logic is coherent in each case. The problem is that coherent logic applied independently to the same investment universe by multiple teams produces correlated portfolios. The differentiation that justifies an active fee — even at the institutional level, and especially at the retail level — lives increasingly in positions ten through forty, not positions one through five.
The case for believing differentiation persists
There is a reasonable counter-argument. Pictet’s robotics mandate gives it genuine exposure to Japanese factory-automation names, European industrial conglomerates and semiconductor design-tool vendors that are structurally absent from a pure-play AI fund. BlueBox’s quality-compounding approach should, in theory, produce a different duration profile than a momentum-driven hyperscaler tilt. BlackRock’s security-software weighting is a thematic bet that AI proliferation increases attack surface — a genuinely separate thesis from owning AI infrastructure directly.
For an investor building a technology allocation from scratch, these distinctions are meaningful. A Pictet robotics position alongside a BNP disruptive technology position is not the same as holding two hyperscaler-tilted funds. The overlap exists, but it does not make the funds identical.
The risks that cut across all of them
The convergence risk, however, is not about normal-market differentiation. It is about tail scenarios. If the hyperscaler capex cycle turns — if cloud providers begin to signal that their AI infrastructure build-out is moderating, or if a credible alternative architecture emerges that disrupts the current silicon hierarchy — the funds in this comparison will not behave as diversifiers of one another. They will fall together, at different speeds, from different peaks, but in the same direction.
That is the honest disclosure that rarely appears prominently enough in fund literature. The mandates are different. The drawdown correlations, in a genuine AI-thesis stress event, would likely not be.
What history suggests about crowded thematic categories
Thematic funds have a consistent pattern in their late-cycle behaviour. As a theme matures and assets accumulate, the eligible universe — as interpreted by institutional consensus — narrows rather than broadens. The funds that were early and genuinely differentiated find their positioning replicated by later entrants with lower conviction and higher AUM. The result is a category that looks diverse on paper and moves as a single trade under pressure.
Whether AI funds are at that stage is genuinely uncertain. The theme is large enough, and the underly ing corporate earnings revision cycle strong enough, that it may not matter for some time. But investors holding two or three funds from this universe in the belief that they are diversifying their AI exposure should look carefully at the actual holdings overlap before drawing that conclusion. The mandates diverge. The top lines may not.
Funds Benchmark provides research and tooling for institutional and private investors. Nothing in this note is investment advice or a recommendation to buy or sell any specific fund. Past performance is not a reliable indicator of future results.