Western Digital adds a $4B buyback as AI servers squeeze memory supply

February 3, 2026
Western Digital adds a $4B buyback as AI servers squeeze memory supply

San Jose, California, Feb 3, 2026, 11:22 (PST)

  • Board authorizes an additional $4.0 billion for share repurchases
  • About $484 million remained under the prior buyback authorization
  • The move comes as AI server demand tightens parts of the memory and storage market

Western Digital said on Tuesday its board approved an additional $4 billion for share repurchases, as demand for its memory chips used in artificial intelligence (AI) servers surges. The stock has risen 57% so far this year after more than tripling in 2025. It had about $484 million left under a prior $2 billion authorization, and the decision comes amid a global shortage that has pushed up memory-chip prices and lead times; last week the company forecast fiscal third-quarter revenue and profit above analysts’ expectations on strong hard drive and flash storage demand from AI servers. (Reuters)

The bigger buyback matters now because it gives the company a lever it can pull fast as AI-driven demand reshapes parts of the storage chain. A repurchase also trims the share count, which can lift earnings per share even if revenue does not move much.

Investors have been looking for signs that AI spending is turning into steady orders for the hardware underneath it. A board-level cash return is one of those signals, even if it is flexible and can be dialed back.

Shares were last up about 4% in late morning trading.

Western Digital announced the new authorization is effective immediately and allows open-market repurchases or private deals, with the pace depending on market conditions. “The expanded $4.0 billion buyback authorization demonstrates our confidence in WD’s future and commitment to shareholder value,” Irving Tan said. The company said it may run repurchases under a Rule 10b5-1 plan — a preset trading plan used to automate purchases under U.S. securities rules — and it can suspend or discontinue the program. (Business Wire)

Western Digital sells hard disk drives and flash storage, which uses NAND, a type of semiconductor memory. Data-center operators and device makers buy those parts to store and move information.

AI servers — the high-end machines used to train and run AI models — generate large data sets that need to be stored, shifted and backed up. That has lifted demand for higher-capacity drives and faster flash storage.

In hard drives, it competes with Seagate Technology; in memory and enterprise storage it runs up against suppliers such as Micron Technology and Samsung Electronics.

A tight memory market has sharpened the battle for supply between AI builders and consumer electronics firms. Even when chipmakers spend more, adding capacity takes time.

But buybacks can slow as fast as they start. If memory prices drop or AI orders cool, Western Digital may choose to keep cash for investment or debt, leaving some of the authorization unused.

Including the $484 million still available under the prior plan, the new approval leaves about $4.5 billion of buyback capacity. Whether that turns into actual purchases will show up in quarterly filings.

2026-02-03 | AI Panic, Buybacks, and Palantir Pops: What to Do Next

Technology News

  • Quantum computing market to grow over 30% CAGR through 2031, fueled by government funding and enterprise adoption
    February 3, 2026, 2:40 PM EST. Astute Analytica projects a global quantum computing market expanding at a CAGR above 30% through 2031. Growth is driven by government funding, enterprise demand, and cloud access through SaaS models that broaden adoption. Data-center workloads in big-data analytics, AI, and simulations push classical systems to the limit, making quantum approaches attractive. The reach spans cybersecurity, healthcare, and finance, prompting collaborations among academia, industry, and national labs to develop scalable quantum architectures. Public-sector backing aims to accelerate research, pilots, and deployment of practical systems. Projections reflect policy support and a push for technology leadership, though outcomes depend on hardware breakthroughs, error correction, and cost trajectories.