Magnetization is nowadays the most frequent cause of inaccuracy in mechanical movements that arrive at the bench. It is not a rare pathology: a smartphone resting on the nightstand, a magnetic clasp on a bag, an MRI scan at the hospital — it takes little to shift a steel balance spring by a few hundredths of a millimeter and push the watch out of tolerance without the wearer noticing immediately.
Magnetic field is measured in Amperes per meter (A/m) or in Gauss: 1 Gauss corresponds to approximately 79.6 A/m. The ISO 764 standard, which defines the minimum threshold to label a watch as antimagnetic, stops at 4,800 A/m (60 Gauss). The problem is that this threshold is far exceeded by everyday objects: a laptop generates up to 16,000 A/m (200 Gauss), and a tablet or smartphone are not far behind. In other words, an ISO 764 certified watch is protected from industrial fields of the 1950s, not from those of 2024.
The historical answer was the Faraday cage: an internal soft iron screen that deflects magnetic flux before it reaches the balance wheel. The Rolex Milgauss, born in 1956 for CERN physicists, and the first IWC Ingenieur use this approach. The cage works, but imposes limits on case thickness and movement visibility.
The modern alternative is to make the movement itself antimagnetic by replacing critical ferromagnetic components with unsusceptible materials: silicon balance springs (Blancpain Fifty Fathoms, caliber 1315), Nivachron (Certina, ETA Powermatic 80), Silinvar (Patek Philippe). Silicon is diamagnetic, does not magnetize, and needs no cage. This allowed METAS to certify the Omega Master Chronometer at 15,000 Gauss, three orders of magnitude beyond ISO 764.
On the bench, diagnosis is simple: bring a compass close to the crystal — if the needle deflects, the movement is magnetized. Demagnetization is done with an alternating current demagnetizer: thirty seconds, zero disassembly, zero risk if executed correctly. What you don't do is pass the watch over a permanent magnet thinking to reverse the process: you make the situation worse.
A special case is the absolute antimagnetic required in EOD and military diving: the Bund ref. 3519 AMAG had to have zero residual field because a diver clearing mines could not afford even a trace of magnetism. That specification is another matter from resistance to environmental fields: it is not enough not to magnetize, you must not already be magnetized at the start.