A rare-earth-free magnet, created by artificial intelligence, could profoundly alter the global balance of technology and geopolitics.
MagNex, developed by the British company Materials Nexus, not only promises competitive performance and a smaller environmental footprint, but also completely dispenses with neodymium and other critical elements, long considered irreplaceable.
Behind the finding, there is no secret laboratory, but an algorithmic platform, based on AI, capable of simulating millions of chemical combinations in record time.
A minor development that could change the world
By: Gabriel E. Levy B.
For decades, the tech industry assumed an unquestionable premise: to build powerful, durable, and compact magnets, it was essential to use rare earths. Neodymium, dysprosium or samarium, elements that are difficult to extract and have a high ecological impact, were central to electric motors, wind generators, audio systems, mobile phones and even guided missiles.
China, which controls more than 60% of the global supply of rare earths and about 90% of their refinement, thus became a key geostrategic player.
In 2010, for example, the Chinese government limited exports of rare earths to Japan, sparking a trade and diplomatic crisis that led many countries to reconsider their dependence on these materials.
In parallel, the European Union, the United States and Japan began to invest in recycling methods, research into substitutes and new extraction processes.
But innovation was progressing slowly. “The development of new magnetic materials without rare earths has historically been very inefficient, based on trial and error,” wrote physicist Karl J. Strnat, a pioneer in the study of rare earth magnets.
Until now, all efforts to find truly viable alternatives had failed in terms of power, durability or scalability. That was the norm… until artificial intelligence arrived.
“We design materials as software is designed”
What sets the MagNex project apart is not only the result, but the method.
Materials Nexus used an artificial intelligence platform to analyze more than 30 million possible combinations of elements, evaluating their electronic structure, magnetization, cost, and ecological footprint.
This process, which under normal conditions would have taken decades, was resolved in a matter of weeks.
AI made it possible not only to reduce the carbon footprint by 70% compared to conventional magnets, but also to reduce production costs by 20%.
As if that were not enough, MagNex can be manufactured entirely with common materials, available in multiple regions of the planet. “We no longer design materials as if we were looking for gold in the mud.
Now we design them as if they were software: accurate, replicable, and optimized from the first code,” said Materials Nexus CEO Jonathan Bean.
This approach responds to an emerging trend in materials science known as Materials Informatics, which combines data mining, machine learning, and computational simulation to accelerate discoveries. According to researcher Ramprasad Rampi of Georgia Tech, “artificial intelligence does not replace the chemist or the physicist, but it offers them a treasure map that simply did not exist before.”
“A silent war for control of critical materials”
The geopolitical background cannot be overlooked. The energy transition and the electrification of transport triggered demand for rare earths, raising prices and international tensions.
In its 2023 report, the International Energy Agency warned that without diversification or substitutes, the world would increasingly rely on a handful of countries for access to strategic materials.
The United States reacted by creating the “Critical Minerals Initiative,” a plan to encourage domestic extraction and sign agreements with allied countries. Europe designed its own “Critical Raw Materials Act”, which seeks to identify reserves and speed up exploitation permits.
But environmentalists and local communities alike have shown resistance to the reopening of mines or the expansion of extractive areas.
The emergence of MagNex may, then, relieve some of this pressure.
Not only because it allows us to circumvent dependence on China, but also because it opens the door to a more sustainable industry, without the radioactive contamination or toxic waste that usually accompany rare earth mining. In the words of economist Jeffrey Sachs, “real change will come not from extracting more, but from needing less.”
However, the transition will not be automatic. Giant companies such as General Motors, Siemens or Tesla still rely heavily on traditional magnets. Changing designs, adapting industrial processes, and certifying new materials takes time, investments, and a new chain of trust.
“From wind power to smartphone: everyone wants fewer rare earths”
The MagNex revolution finds its testing ground in everyday applications.
In the automotive industry, for example, electric motors in hybrid and all-electric vehicles use rare-earth magnets to optimize performance and reduce motor size.
A Toyota Prius contains more than 1 kg of neodymium and dysprosium. Replacing these elements with MagNex, without loss of power, can represent a massive change in costs and logistics.
In the wind sector, each turbine can contain between 300 and 600 kg of permanent magnets. New offshore wind farms in the North Sea or on the Atlantic coast of the United States are highly dependent on the supply of rare earths.
Energy companies such as Ørsted and Iberdrola are already evaluating alternative materials, aware that an interruption in supply could delay million-dollar projects.
The same is true for consumer electronics: from noise-canceling headphones to hard drives or smartphones.
Apple, which in 2021 promised to “phase out its dependence on critical materials”, could find in MagNex a concrete, industrializable and more ethical solution.
Even in national security sectors, such as defense, replacing critical magnets can have strategic implications.
According to the Pentagon, more than 90% of the magnetic components used in radars, missiles or drones originate outside the United States. The possibility of making local magnets from abundant materials can alter this structural dependence.
MagNex isn’t alone.
Other companies, such as Japan’s Daido Steel and North America’s Niron Magnetics, are also developing rare-earth-free magnets. However, the advantage of Materials Nexus lies in its AI platform, which can adapt to new challenges, such as creating alloys for extreme temperatures, corrosive conditions, or variable magnetic fields.
In that race, software can end up being just as valuable as hardware.
In conclusion, MagNex represents much more than a new type of magnet: it is the symbol of an industrial transition where artificial intelligence, sustainability and strategic autonomy converge. By eliminating the need for rare earths, it not only alleviates ecological and geopolitical pressures, but inaugurates a new way of designing materials, from the code and not from the mine.
References:
- Strnat, K. J. (1988). Rare Earth Cobalt Permanent Magnets. IEEE Transactions on Magnetics.
- Ramprasad, R., Batra, R., Pilania, G., Mannodi-Kanakkithodi, A., & Kim, C. (2017). Machine learning in materials informatics: recent applications and prospects. npj Computational Materials.
- International Energy Agency (IEA), The Role of Critical Minerals in Clean Energy Transitions, 2023.
- Sachs, J. D. (2022). The Age of Sustainable Development. Columbia University Press.
