Trusting Autonomous Vehicles After a Drink? It's Still Drunk Driving Even Without Hands on the Wheel

Published 01 Apr.2026 15:30(KST)

Does operating a vehicle after drinking, while using autonomous driving features, constitute drunk driving? Although autonomous driving technology is advancing rapidly, the current legal framework still maintains the premise that the "driver is a human." Court rulings also hold that the driver retains responsibility, even when autonomous driving functions are engaged.

Recently, Tesla has been experiencing rapid growth in the domestic electric vehicle market. As of 2025, sales approached 60,000 units, more than doubling compared to the previous year. In November 2025, Tesla introduced the Full-Self Driving Supervised (FSD) system in Korea. As autonomous driving technology becomes more widespread, the legal issues surrounding it are once again drawing attention.

An image to help understand the article. Pixabay

Tesla Vehicles Sold in Korea Offer Partial Autonomous Driving

The current Road Traffic Act clearly outlines the driver's responsibilities even during autonomous driving scenarios. The FSD in Tesla vehicles sold domestically is classified as partial automation (Level 2) under the Autonomous Vehicle Act, rather than "fully autonomous driving," which would require no human intervention. At this stage, the driver must always keep their eyes on the road and be ready to intervene immediately if needed.

Therefore, even if autonomous driving features are activated, the act of moving the vehicle itself is still legally considered driving. Experts note that if this is done while intoxicated, it is evaluated in the same way as conventional drunk driving.

Geunwoo Lee (Judicial Research and Training Institute, 35th class), attorney at HwaWoo, stated, "While the Supreme Court includes a purposive element in the concept of driving, if a driver intentionally activates autonomous driving while intoxicated, this can sufficiently be regarded as intentional driving," adding, "It is difficult to argue for reduced responsibility solely because autonomous mode was used."

Kang Kyunghee (34th class), attorney at MNL and Vice President of the Korea Automobile Mobility Safety Association, also explained, "Currently, Tesla's FSD is a technology that requires continuous driver intervention," and, "If you operate a vehicle after drinking by enabling autonomous driving, it is reasonable to consider this drunk driving under the Road Traffic Act."

Court rulings have reached similar conclusions. The Ansan branch of the Suwon District Court sentenced a defendant to one year in prison, suspended for two years, for driving approximately 46 kilometers using Tesla's autonomous driving feature with a blood alcohol concentration of 0.154% (2021Godan2397). The Seoul Southern District Court also imposed a fine in a case where a driver caused an accident while using autonomous driving features under the influence (2020Godan1100).

Can Autonomous Driving Data Be Used as Evidence for Drunk Driving?

Some analysts suggest that autonomous driving technology may actually make it easier to prove drunk driving. Vehicles are equipped with accident data recorders (EDR), data storage systems for autonomous driving information (DSSAD), and driver monitoring systems (DMS), all of which objectively record the timing and nature of driver interventions. Video data collected during autonomous driving may also be used to reconstruct the circumstances at the time of an accident.

Dongyoung Song (39th class), attorney at Jipyung, stated, "There are no explicit regulations regarding the use of vehicle log records as evidence in civil or criminal cases, but if law enforcement obtains them through voluntary submission or proper legal procedures, they can be used as evidence," adding, "In some cases, such data could even be used to prove that the driver was not drunk driving."

Attorney Geunwoo Lee, assuming that legal procedures for law enforcement to access Tesla vehicle data are established, analyzed, "FSD log data can serve as evidence supplementing or replacing traditional investigative techniques regarding delayed driver intervention or abnormal operating patterns. Since the data is transmitted and stored in real time on Tesla's servers, drivers cannot manipulate the system records after the fact, which increases the evidentiary value."