Jennison Associates LLC Rewires Its AI Bet Around Infrastructure

Jennison’s biggest buys are not random stock picks; they cluster around the physical and logical rails of the AI economy. The portfolio tells a clear story: they want to own the gear that feeds the data center arms race, the networks that route that traffic, and a handful of platforms that can monetize AI-native workflows.

• Palantir: A fresh $1.57B position at 1.08% weight, despite being about -20.2% below their average entry, is a bold statement. Jennison is willing to sit through volatility to own a software vendor with high strategic leverage to defense, intelligence and data-driven operations, even as near-term sentiment sours.
• KLA and Lam Research: A new $566.1M KLA stake plus a near-3x Lam Research add (LRCX up 296.9% by shares, +$539.1M) is classic “sell the hype, buy the shovels” behavior. They are compounding an already large semiconductor exposure by moving further up the capex stack into process control and wafer fab equipment.
• Cloudflare and Broadcom: A 135.8% share increase in Cloudflare (+$522.2M) and incremental buying in Broadcom (+$275.7M) show a tilt toward bandwidth, security and custom silicon as AI workloads explode. These are leveraged plays on traffic growth and on-prem/cloud AI deployments rather than on any single model winner.
• GE Aerospace, GE Vernova, Mastercard, Airbnb, TJX, Merck and Texas Instruments: Adds across these names — from travel capacity and energy transition to payments and analog semis — show Jennison selectively broadening from pure software into real-economy beneficiaries of a long expansion, but only where they see durable moats and pricing power.

Funding these moves required real pruning, not cosmetic rebalancing. The heaviest trims line up almost perfectly with the parts of the growth complex that look crowded, fully valued, or structurally less advantaged in the next leg of the cycle.

• Megacap platform trims: Microsoft (-15.0% shares, -$1.17B), Nvidia (-5.2%, -$710.5M), Alphabet (GOOGL -9.1%, -$573.8M; GOOG modestly lower), Apple (-5.8%, -$527.8M), Amazon (-4.4%, -$401.9M) and Meta (-5.8%, -$277.3M) are all still core, but they’re now the ATM. With enormous embedded gains — Nvidia is up over 2,100% versus their average buy — Jennison is recycling capital from consensus AI winners into higher-beta, under-owned infrastructure.
• High-multiple cloud software: Snowflake (-38.3% shares, -$669.5M), CrowdStrike (-30.5%, -$655.4M), Cadence (-25.3%, -$542.9M), Oracle (-26.3%, -$364.7M), Datadog (-22.3%, -$200.8M), AppLovin (-15.9%) and Cloudflare last quarter (before this quarter’s add) all show up as material sources of cash. Note that Snowflake and CrowdStrike are trimmed even with relatively modest or negative gains versus average cost — this looks more like a quality upgrade than simple profit-taking.
• Consumer and media: Netflix (-19.7%, -$661.1M), Disney (-17.1%), O’Reilly (-22.7%), Shopify (-14.1%), MercadoLibre (-13.8%), Vertex (-15.7%) and a range of smaller discretionary names are being leaned out. Jennison is clearly less interested in advertising and discretionary spending sensitivity, and more focused on transaction fees, travel capacity and off-price retail where they did add (Airbnb, TJX).

This is not a de-grossing quarter. It is a rotation: cash is coming out of crowded, fully harvested winners and fragile growth to finance a more infrastructure-heavy, idiosyncratic growth book.

On the surface, sector weights look static. Technology barely budged — 61.23% to 61.09% — and Consumer Discretionary, Health Care, Finance and Utilities each moved by less than 0.3%. But the composition of those buckets is quietly changing in ways that matter.

Inside Technology, Jennison is shifting from megacap platforms and application software toward semiconductors, equipment and networking. The new KLA stake, the tripling of Lam Research, and incremental increases in Broadcom, Texas Instruments and GE Aerospace all lean into the hardware and industrial side of AI and compute, even as Snowflake, Oracle, CrowdStrike and Datadog get cut back.

Consumer Discretionary remains a sizeable 18.0%-ish sleeve, but it’s being refocused. Trims to Netflix, Disney, O’Reilly and others contrast with outsized adds to Airbnb (+9,331.0% by shares, +$396.4M) and TJX (+82.8% shares, +$238.6M), pushing the book toward travel, experiences and value retail instead of pure streaming and autos.

Real Estate (where Mastercard, Visa and MercadoLibre are slotted in this dataset) inches up from 4.11% to 4.34% as Mastercard is increased. That effectively means more exposure to transaction volumes and digital commerce. Meanwhile, the small but growing “unclassified” bucket is really GE Vernova — a levered play on grid and energy infrastructure — underscoring a broader theme: Jennison wants to own the pipes and power behind digital demand.

Taken together, this quarter says Jennison believes the AI and digitalization cycle is early in its capex phase, even if some front-line beneficiaries look fully priced. They are intentionally swapping a slice of low-volatility megacap exposure for more cyclical, but higher-upside, infrastructure and tools.

By adding aggressively to semicap names, networking, aerospace and energy transition plays, Jennison is positioning for years of elevated spending on data centers, manufacturing and travel. The new Palantir position suggests they still want upside to AI-native software workflows, but they prefer firms with deep ties into government and mission-critical analytics instead of broad horizontal SaaS.

At the same time, incremental increases in payments and marketplaces, together with trims in ad-driven media and generic cloud software, point to a preference for transaction-linked, cash-generative models over eyeball-count stories. If growth slows or rates stay higher, they are betting that fee-takers, toll collectors and infrastructure owners will hold up better than the average “AI beneficiary.”

None of this is riskless; the portfolio is still heavily concentrated, with the top 10 at 45.9% and Technology over 60%. But after a -11.36% quarter, Jennison’s response is unambiguous: lean further into the structural AI buildout, just with more emphasis on shovels, pipes and runways than on the marquee names that led the last leg.