|
Getting your Trinity Audio player ready...
|
The true power in artificial intelligence (AI) has shifted from model innovation to the control of infrastructure, capital, and platform ecosystems, according to the Artificial Power: 2025 Landscape Report by the AI Now Institute.
The report revealed that the global AI race is no longer driven primarily by breakthroughs in large language models or futuristic visions of general intelligence. Instead, dominance now lies with the organisations that own and control the compute, data, and distribution networks powering AI deployment.
“This isn’t just about building smarter models; it’s about who owns the mechanisms of production and distribution,” the report stated.
The findings show that companies investing heavily in AI infrastructure, such as massive data centres, specialised chips, and global cloud networks, are consolidating long-term power.
These investments have become strategic assets that set the ‘rules of the game,’ granting firms enduring advantages beyond short-term technological leaps.
Read also: The first significant mass-scale artificial intelligence (AI) deception case
The report described this growing dominance as a ‘Too Big to Fail’ scenario, where a handful of infrastructure-rich AI platform owners are emerging as new global gatekeepers.
For businesses and users, it noted that the value of AI now lies less in owning models and more in controlling the surrounding ecosystem. Companies that embed AI deeply into their data pipelines, operations, and digital platforms will capture more value and more influence than those that merely deploy AI tools.
Speaking about policy, the report warned that society is increasingly being shaped by AI platforms rather than merely using them. This concentration of infrastructure and capital raises urgent concerns about fairness, access, transparency, and control.
For Africa and other emerging regions, the report urged stakeholders to focus on building the foundational layers of AI, which are data infrastructure, access frameworks, and distribution capacity, if they hope to convert AI innovation into widespread impact.
“It’s not just about having cool models. It’s about having the infrastructure and ecosystem to turn those models into real economic and social power,” the report stated.
