Digitalization of finance has rapidly transformed the global financial system by introducing new complex technologies. As digital finance grows more interconnected, we show in this work that solar geomagnetic storms, i.e., an exogenous space-weather shock, pose a systemic risk comparable to climate transition pressures, with measurable consequences for Bitcoin (BTC) mining. BTC mining, an energy-intensive process reliant on stable electricity and continuous network connectivity, faces increasing risks from solar storms, i.e., geomagnetic disturbances, capable of disrupting power grids and digital infrastructure. This study introduces a novel step- by-step ANN-based causality test, leveraging Artificial Neural Networks (ANNs) to examine the nonlinear and dynamic causal relationship between solar storms and BTC mining activity. In this context, we formally establish consistency, asymptotic normality and other useful properties of the proposed ANN estimator. Using daily data from 09/2014 to 01/2023, our multi-horizon analysis identifies statistically significant short-, medium-, and long-term effects, revealing that severe solar storms immediately destabilize networks, delaying transaction confirmations and reducing BTC activity. Furthermore, we quantify the economic significance of these disruptions. During 2015 alone, geomagnetic storms were associated with a cumulative reduction of nearly 3 million BTC in trading activity, equivalent to an estimated value of around 883 million. The findings position solar storms as a critical but overlooked risk factor in digital finance, with implications for cryptocurrency market stability, mining infrastructure resilience, and regulatory oversight. Our research stresses the need for risk monitoring, geographically diversified mining operations, and enhanced grid stability measures.

Keywords: Digital Finance, Bitcoin, AI, Artificial Neural Networks, Solar Storms.

JEL Classification: C22, C50, C51, C58.