Researchers use artificial intelligence to identify subtle “slow slip” activity beneath California’s San Andreas Fault that earlier methods did not detect. Slow-slip events are low-velocity movements on fault planes that can release built-up tectonic stress over hours, days, or longer, without producing the clear, rapid shaking associated with conventional earthquakes. According to the reporting, the AI analysis finds additional signals consistent with these silent fault adjustments. The studies also describe changes in low-frequency earthquake activity following the slow-slip episodes, suggesting a connection between the silent movements and later seismic behavior. While the outlets discuss the potential for improving understanding and possibly earthquake forecasting, they present the work primarily as a detection and interpretation effort: it highlights previously missed fault behavior and examines how that behavior relates to subsequent seismic measurements. The findings therefore focus on what the AI detects and how the detected events correlate with other observed seismic signals, rather than claiming a direct, immediate ability to predict the timing or magnitude of a future major earthquake.