Zuckerberg Wants AI for Everyone — But Who Really Benefits?

Picture this: a teenager in Nairobi, a student in Lagos, a young developer in Accra — all running powerful artificial intelligence directly from their laptops, no cloud subscription, no corporate gatekeeping, no monthly fee eating into their data budget. That is the future Mark Zuckerberg says he is building. On Monday, the Meta CEO made his boldest move yet in the global AI race, unveiling a new open-source model designed to run on personal computers and publishing a sweeping 6,500-word essay titled “The Future is for Everyone” — a phrase that lands very differently depending on where in the world you are reading it.

The centerpiece of the announcement is Muse Glimmer, a leaner AI model built to run directly on a Mac or PC equipped with a single graphics card, handling what the industry calls agentic tasks — actions where AI operates with some degree of autonomy to complete goals on your behalf. Alongside it, Meta confirmed it would open the weights of Muse Spark 1.2, its most capable model to date, making it freely downloadable for anyone on the planet. Weights, in plain terms, are the internal rules and calculations that shape how an AI system thinks and responds — and releasing them is a big deal, because it means developers anywhere can inspect, adapt and build on Meta’s work without asking permission.

Zuckerberg’s essay is not a press release. It reads more like a manifesto, and it is one clearly aimed at winning a political and ideological argument as much as a technological one. His main rivals, OpenAI and Anthropic, operate largely closed systems — meaning their core models stay locked behind APIs and paywalls, accessible only on their terms. Zuckerberg frames this as a concentration of power, warning that building AI “for companies, governments or other institutions” tips the balance away from ordinary people. “Rather than centralizing superintelligence,” he writes, “we should distribute it widely and give every person the ability to direct it.” It is a compelling line. Whether it holds up under scrutiny is another question entirely.

The backdrop to all of this is a United States government growing increasingly nervous about artificial intelligence and who controls it. In June, US President Donald Trump signed an executive order establishing a voluntary federal review of national security risks before advanced AI models can be released — a move that signals Washington is watching the sector closely, even if it has not yet moved to regulate it hard. Zuckerberg used his essay to push back, arguing that placing extra restrictions on American AI companies only hands an advantage to foreign competitors, particularly Chinese labs. He pointed specifically to Moonshot and DeepSeek, two Chinese AI developers whose models are also open-weight, as evidence that openness is already a global norm — and that the US risks falling behind if it tightens the screws on its own industry.

The geopolitical framing is deliberate, and it works on multiple levels. For a young African audience watching this debate unfold, the China-versus-America narrative can feel distant — but the underlying question of who controls powerful technology is anything but. Most of the AI tools shaping daily digital life across the continent are built, owned and governed in the Global North, with African users generating data that feeds these systems while having little say in how they are designed or deployed. Zuckerberg’s open-source pitch, at least in theory, chips away at that dynamic by allowing local developers to take Meta’s models and build something genuinely adapted to their own contexts, languages and needs.

Still, openness has its limits, and the debate inside the AI research community is sharper than Zuckerberg’s essay lets on. Some experts argue that open-weight models consistently trail the most advanced closed systems in raw performance — meaning the cutting-edge capabilities tend to remain locked behind the doors of well-funded labs regardless. Meta’s own governance announcement reflects this tension: Zuckerberg said the company would establish a structure giving its independent directors the power to approve safety criteria before models are released, a concession that even open AI needs guardrails. The promise of on-device AI — faster, cheaper, private, accessible without a reliable internet connection — is real and meaningful, especially in regions where connectivity remains expensive and patchy. But the gap between a model that runs on a MacBook and one that genuinely serves a first-time smartphone user in rural Mozambique remains vast.

What Zuckerberg has done, whatever one thinks of his motives, is force a conversation that matters. The question of whether artificial intelligence becomes a tool of liberation or a new layer of digital dependency is not abstract — it is being answered right now, in the design choices, licensing terms and political lobbying happening in Silicon Valley and Beijing. For the generation growing up with AI woven into their phones, their schoolwork and their economic futures, understanding who is making those choices — and why — is not optional. The future, Zuckerberg insists, is for everyone. The work of making that actually true is still very much unfinished.