Artificial intelligence has undeniably transformed productivity and daily life, but its development has also concentrated power in the hands of a few corporations. Giants such as Google (Gemini), OpenAI (ChatGPT), X (Grok), and Anthropic (Claude) dominate the ecosystem, holding most of the computing resources, data, and top talent.
This centralisation raises concerns about bias, privacy, and the unchecked influence of private firms over technologies that increasingly shape society.
Critics argue that centralised AI models collect and monetise vast amounts of personal and corporate data with little transparency.
A Stanford University study in 2025 found users perceive large language models to lean politically left, while controversies have emerged around Grok allegedly producing antisemitic rhetoric and Gemini misrepresenting historical figures.
Beyond bias, scaling constraints are evident, data centres already strain global electricity use and are projected to consume 20% of global power by 2030. Centralised systems also create single points of failure, making them attractive targets for hackers.
In response, interest in decentralised AI is accelerating. Valued at $550.7 million in 2024, the sector is expected to reach $4.33 billion by 2034. Unlike traditional models, decentralised systems keep raw data on local devices, sharing only trained insights across networks secured by blockchain.
This distributes control among participants rather than concentrating it with a single company.
The benefits are compelling. Individuals and organisations retain control over their data, deciding what to share. Training across a decentralised network introduces greater diversity of perspectives, reducing systemic bias.
By distributing computation across devices, the model scales efficiently without relying on energy-hungry central servers. Security also improves without a central point of attack, blockchain adds resilience while much sensitive data never leaves the user’s device.
Advocates link this shift back to early cypherpunk ideals.
As Eric Hughes wrote in A Cypherpunk’s Manifesto, cryptography was meant to safeguard privacy and liberty in the digital age. While cryptocurrencies drifted toward profit-seeking, decentralised AI could represent a return to those original principles including rebalancing power, protecting privacy, and building a more sustainable digital future.
