Meta is stepping up its push into physical retail, open-source AI, and military-grade AR and VR technology, but each move is attracting scrutiny and raising new questions about ethics, transparency, and the company’s strategic direction.
Physical Stores Selling Smart Glasses
Meta is reportedly preparing to open a new wave of physical retail stores in a strategic bid to push sales of its Ray-Ban smart glasses and other wearable devices. The move (first revealed by Business Insider) signals a move from virtual ambitions into tangible retail expansion as the company looks to solidify its position in the emerging face-computing market.
It’s worth noting here that this won’t be Meta’s first foray into bricks-and-mortar. For example, the company launched its debut physical store in Burlingame, California in 2022, followed by a pop-up in Los Angeles. But this latest round of hiring and planning suggests a much broader rollout is on the cards.
The logic behind Meta’s planned move appears to be that smart glasses, especially those blending AR features with fashion, are inherently tactile. Trying them on in person can make or break a sale. It seems that although Meta reportedly sold over 1 million Ray-Ban Meta smart glasses in 2024 alone, CEO Mark Zuckerberg is said to have challenged staff to raise that figure to 5 million units, prompting the search for a more immersive retail strategy.
In-Store Demos
In-store demos will also likely help Meta showcase its Meta Quest VR headsets, especially as rivals like Apple raise the stakes with more premium offerings like the Vision Pro, priced at $3,499 but struggling to attract mass-market adoption.
By creating dedicated physical spaces, Meta seems to believe it could address two problems at once, i.e. differentiating its devices from commodity tech, and humanising a brand that’s often criticised for being too virtual and too data-hungry.
For business users, this matters. For example, the rise of face-worn computing is set to impact fields from healthcare to logistics. With AI-assisted smart glasses already capable of real-time transcription, photo capture, and even livestreaming, the line between personal wearables and professional tools is blurring fast.
However, it seems that retail has proven a tricky terrain for Big Tech. For example, Microsoft famously shuttered its 83 stores in 2020, and Amazon has scaled back its ambitions after mixed success with physical shops. Meta’s challenge, therefore, may be to offer something more experiential than transactional, and something that convinces users and developers alike that these devices are more than gadgets.
“Open Washing” Accusations Cloud Meta’s AI Open Source Push
At the same time that Meta is championing openness in AI, the company is facing renewed criticism for allegedly misrepresenting the nature of its flagship Llama models, with critics accusing it of “open washing.”
This latest controversy stems from Meta’s role in sponsoring a Linux Foundation research paper, The Economic and Workforce Impacts of Open Source AI, published in May. The report highlights the cost savings and innovation benefits of open source AI (OSAI), noting that 89 per cent of AI-adopting organisations use some form of open source infrastructure, and that open models are significantly cheaper to deploy than proprietary ones, findings that are hard to ignore for small businesses and tech start-ups.
Meta’s involvement in the report, however, has triggered backlash. For example, critics, including OpenUK CEO Amanda Brock, argue that Llama does not meet the widely accepted Open Source Definition (OSD), largely due to commercial use restrictions embedded in its licence.
“Llama isn’t ‘open source’, whatever definition you choose to use for open source,” Brock stated. “We rely on open source being usable by anyone for any purpose, and Llama is not.”
The nuance here is key. Meta’s Llama models (including Llama 2 and 3) are open access, meaning researchers and developers can use them freely in many cases. However, the restrictions on high-scale commercial use mean they fall short of being truly “open source” under OSI standards.
For Meta, the implications are twofold. First, its marketing message around openness risks losing credibility, especially as regulators in the EU and US begin using “open source” as a basis for liability exceptions in AI laws. Second, it could jeopardise Meta’s appeal to developers who value transparency, forkability, and independence from Big Tech.
The Linux Foundation report itself finds that open models are being adopted more heavily by small businesses than large enterprises, with smaller firms citing lower costs and greater flexibility as primary drivers. If Llama isn’t genuinely open, these businesses could end up relying on what they believe is community-driven infrastructure, only to face legal grey areas or cost barriers later.
While Meta has made real contributions to the open AI ecosystem, including the release of PyTorch and participation in Hugging Face, it’s likely that the wider industry is watching closely to see whether the company’s vision of openness is consistent, or just convenient.
From Metaverse to Military
In yet another Meta development, and this time one that caught many observers by surprise, Meta has signed a deal with Anduril Industries, a fast-growing US defence contractor, to build AR and VR devices for military use.
The irony hasn’t gone unnoticed. For example, Anduril’s founder, Palmer Luckey, was famously ousted from Meta’s predecessor, Facebook, back in 2017, reportedly over political donations and internal disagreements. Now, Luckey is working with his former employer on a project aimed at turning soldiers into what he describes as “technomancers.”
“I am glad to be working with Meta once again,” said Luckey in a statement. “The products we are building with Meta do just that.”
Battlefield-Ready Technology
According to Meta, the partnership will leverage its expertise in AI and extended reality (XR) to deliver battlefield-ready technology that enhances real-time situational awareness. The systems are expected to integrate with Anduril’s Lattice platform, a command-and-control interface powered by AI that overlays live battlefield intelligence into soldiers’ fields of view.
Could Actually Make Money
Meta’s AR and VR ambitions have so far been expensive and, arguably, unproven. Its Reality Labs division lost $4.2 billion in Q1 2025, part of a long-term investment strategy that has seen Meta burn through over $80 billion on immersive tech since acquiring Oculus in 2014. However, with the US military’s proposed $1 trillion budget, the defence market could finally offer a return.
Progress and Pitfalls
For business users and the wider XR market, the defence partnership signals both progress and potential pitfalls. For example, military involvement could fast-track innovation, enhance hardware capabilities, and make advanced technologies more accessible for commercial use. However, it also raises serious questions about Meta’s role in surveillance, data handling, and the broader ethical implications of merging consumer tech with military objectives.
Notably, Microsoft handed off its own US Army IVAS contract to Anduril earlier this year, having struggled to deliver effective headsets with its now-discontinued HoloLens. That leaves Meta and Anduril in what appears to be a strong position to lead the next wave of military-grade XR, and potentially commercial spin-offs.
What Does This Mean For Your Business?
Taken together, Meta’s latest moves suggest a company trying to reposition itself at the centre of three major battlegrounds: consumer hardware, AI ethics and extended reality. Each initiative carries its own rationale. Retail stores aim to offer hands-on experience, the push for AI openness suggests leadership in cost-efficient tools, and the defence partnership seeks a path to long-term commercial viability. However, the connections between these efforts reveal a deeper tension. Is Meta attempting to cater to every audience at once, or is it spreading its efforts too widely in a fast-changing technological landscape?
The physical retail expansion is perhaps the most straightforward. It gives Meta a tangible way to demonstrate products that have, until now, lived mostly in speculative hype cycles. Smart glasses, in particular, are on the verge of moving from novelty to utility, and a physical showroom model could help convert curiosity into confidence. For UK businesses operating in sectors like healthcare, manufacturing, logistics or retail, that’s potentially game-changing. If the technology works and is easy to trial and adopt, it could speed up the mainstreaming of AR in workplace settings.
However, it’s the open-source AI row that cuts to the heart of Meta’s credibility. The company is trying to paint itself as a champion of openness, cost savings, and accessibility, which is a narrative that appeals to developers and small firms alike, but the reality of Llama’s licensing restrictions muddies that message. If Meta is seen to be overstating its openness, or using community narratives to mask corporate control, it could backfire with the very audiences it’s hoping to win over. For UK tech start-ups and SMEs, who often rely on open source to compete with bigger players, the difference between “open” and “open enough” represents a business risk.
The Anduril partnership adds another layer of complexity. On paper, it could finally make Meta’s multibillion-dollar investment in XR technologies pay off. But aligning with military objectives also risks alienating consumers, employees, and partners who are wary of how immersive tech might be used in surveillance or combat. In a world increasingly conscious of tech’s societal impacts, even commercial buyers may start asking harder questions about the provenance and purpose of the tools they deploy.
From businesses and developers to policymakers, the message appears to be that Meta is doubling down on its hardware and AI bets. That said, how it navigates trust, ethics and transparency will shape not only its own future, but the broader acceptance of emerging tech across the board. What comes next may, therefore, depend less on product specs and more on public perception, legal scrutiny, and whether Meta can balance innovation with genuine accountability.