Epilepsy is a common chronic neurological disorder. Drug-resistant epilepsy (DRE) is defined as failure to achieve sustained seizure freedom after adequate trials of two appropriately chosen and tolerated anti-seizure medications (ASMs) regimens (as monotherapies or in combination). In this population, surgical resection and neuromodulation can effectively reduce seizure burden; however, their effectiveness is limited by uncertainty in epileptogenic zone localization, interindividual variability in stimulation targets and parameters, and procedure-related risks. As an emerging technology, brain–computer interface (BCI) offer a closed-loop framework of real-time monitoring, state decoding, decision-making, and intervention delivery, shifting therapy from fixed-parameter open-loop approaches to biomarker-driven, dynamically personalized modulation and providing new avenues for precision, individualized treatment of DRE. This review summarizes signal acquisition and decoding strategies for epilepsy BCI, closed-loop intervention paradigms, and key challenges in clinical translation.