Has Allbirds Thought Through Its AI Bet — Or Is This A Leap Too Far?

Allbirds, which was launched in 2016 with wool-based running shoes, managed to amass a devoted following among tech-savvy, environmentally conscious buyers. So much so that the company went public on the Nasdaq in November 2021 and raised roughly $303 million at a valuation near $4.1 billion. Everything looked good on paper, but what followed was a long, slow erosion of both revenue and investor confidence.

Allbirds' sustainability credentials it had built over the years and its cultural cachet never translated into durable profits in a brutally competitive retail market. The company had closed all of its full-price U.S. retail stores in February 2026. The revenue had fallen nearly 50 per cent from peak levels by the time the unexpected AI pivot was announced. The $39 million sale price was a clear indication of how far things had fallen since the IPO.

The decision to exit shoes entirely instead of restructuring or finding a new partner within footwear spectrum reflected a stark conclusion: no viable path remained in the original market. So leadership looked elsewhere and said it found a gap in the AI compute supply chain.

NewBird AI's GPU Strategy and What It Actually Involves

Allbirds' plan is to focus on the market for graphics processing unit (GPU) rental. These are high-powered chips that drive AI model training and inference work. The company managed to secure $50 million in funding to acquire GPU assets, according to The Washington Times, with that came the announcement of rebranding and the footwear sale.

The news of rebranding, from Allbirds to NewBird AI, skyrocketed its share price by more than 580 per cent. That may sound like a lot, but the company's stock market value had plummeted from $500 per share to around $2.50 just before the AI pivot announcement.

Data center vacancy rates have fallen to historic lows, which means businesses seeking compute capacity cannot always get it through the dominant cloud providers, sometimes called hyperscalers, such as Amazon Web Services, Microsoft Azure, and Google Cloud. This is where NewBird AI sees a gap, which intends to offer long-term GPU lease arrangements to customers who cannot get what they need from those larger players.

"The rise of AI development and adoption has created unprecedented structural demand for specialized, high-performance compute that the market is struggling to meet," the company said in a statement. "NewBird AI is being built to help close that gap."

Is it an entirely novel idea? Not exactly. Companies like CoreWeave and Lambda Labs already operate in the GPU-as-a-service space, with years of infrastructure experience and established client relationships behind them.

Industry experts have also expressed skepticism about NewBird AI's ability to compete meaningfully in AI infrastructure, even as markets rewarded the announcement as the share price went from $3 to $17.

Building a credible GPU rental operation requires not just capital to buy chips, but deep expertise in data center operations, networking, cooling, and enterprise sales. Allbirds has no track record of any of it.

The $50 million is a meaningful sum for a company of Allbirds' recent size, but it is not much when compared to rivals that have raised billions. NewBird AI has not publicly said which GPU models it plans to buy or the kind of data center facilities it will use, or which customers it has lined up.

For the sustainability-minded consumers in Silicon Valley who made Allbirds a cultural moment in the late 2010s, the rebrand translates to a complete departure from the company's founding mission. Tim Brown and Joey Zwillinger built the brand around the argument that profit and environmental responsibility were compatible. But then selling that brand for $39 million to a consumer goods holding company, and pointing the corporate shell toward GPU racks, closes that chapter without any closure.