BlackRock Names Three AI-Focused Stocks Among Top Picks

BlackRock has named Astera Labs (ALAB), Celestica (CLS), and Arista Networks (ANET) among the AI-focused stocks it considers most important right now, placing all three inside a broader list drawn from the holdings of its iShares Future AI & Tech ETF (ARTY). The three companies sit outside the mega-cap AI names that dominate most headlines: semiconductor connectivity, contract electronics manufacturing, and networking equipment respectively, but BlackRock's framing puts them at the center of the physical buildout of AI data-center infrastructure rather than at the model or application layer.

BlackRock's broader AI research this year has centered on the idea that the AI buildout is increasingly gated by supply-side bottlenecks, power, chips, skilled labor, and specialized components, rather than by demand, and its list of important AI stocks appears to be built around companies positioned to benefit from those bottlenecks rather than from the AI applications sitting on top of them. That lens favors infrastructure and picks-and-shovels businesses: companies that supply the interconnects, manufacturing capacity, and networking gear that hyperscalers and chipmakers depend on to actually deploy AI compute at scale, regardless of which model or application ultimately wins.

Within that framing, each company plays a distinct, well-understood role in AI infrastructure. ALAB makes connectivity silicon, PCIe retimers, CXL memory controllers, and related interconnect chips, that sit inside AI server racks to keep high-speed signals intact as GPU clusters scale up. CLS operates further down the supply chain as a contract manufacturer, assembling and testing the custom server and networking hardware that hyperscalers design in-house but don't build themselves. ANET supplies high-capacity networking switches used to link GPUs together inside AI clusters, a category where bandwidth and low latency are directly tied to how efficiently large training and inference jobs can run. The available reporting does not lay out BlackRock's precise internal reasoning stock-by-stock, but the shared thread across all three is exposure to the physical infrastructure layer of AI rather than to AI software or foundation models directly.

For investors, being named among BlackRock's AI picks is a signal of thematic relevance rather than a price target or formal recommendation; index and ETF-driven flows into stocks like these can amplify moves in either direction without changing the underlying fundamentals. All three names have already re-rated substantially on AI-infrastructure demand, which raises the bar for continued outperformance and makes them more sensitive to any slowdown in hyperscaler capital spending, delays in data-center buildouts, or supply constraints in components like power and cooling. The inclusion is nonetheless notable: it puts a large asset manager's research effort behind the thesis that the picks-and-shovels layer of AI, not just the chipmakers and hyperscalers most investors already watch, deserves a place in a diversified AI allocation.

Related Stocks

• BLK
• ANET
• AVGO
• PLTR