In a recent Reuters report, Germany’s Federal Office for Motor Vehicles (KBA) said it would not have approved the Autopilot system installed on Tesla cars if the technology was still in a “beta-phase” version. Following this, Tesla CEO Elon Musk has now come out to clarify the carmaker’s use of the term “beta.”

On his official Twitter page, Musk explained that the term “beta” is not used in the standard sense, but rather an explicit warning to make sure that drivers “don’t get comfortable” with the system i.e. sleeping in the car with Autopilot engaged. Put simply, drivers have to be at the ready to take over if needed.

According to Musk, Tesla’s Autopilot system would need to notch up at least one billion miles of real-world driving before it can leave the “beta” phase. He added that even though Tesla had performed extensive Autopilot testing on its fleet, “there is no substitute for real world experience.”

“With less than one billion miles, there simply isn’t enough data. One billion is a necessary but not necessarily sufficient condition,” he tweeted. When asked when the company will hit the billion-mile mark, Musk expects to hit said target in around six months, after which all Autopilot-equipped Teslas will be offered a software update that includes “hundreds of refinements to handle rare corner cases.”

Back in May, Tesla’s Director of Autopilot Programs, Sterling Anderson confirmed that the company has collected 780 million miles worth of Autopilot data and 100 million miles of active Autopilot use, according to a report by Electrek. He added that the company is adding a million more miles every 10 hours – thanks to a fleet of over 70,000 Teslas equipped with Autopilot hardware.

Tesla’s Autopilot system has come under intense scrutiny from the US National Highway Traffic Safety Administration (NHTSA) following the first-known fatality involving the system. The “self-driving” feature has also been linked with two Tesla Model X crashes in Pennsylvania and Montana, recently.