AMCI Testing reported that its drivers intervened more than 75 times during a drive of over 1,000 miles in a 2024 Tesla Model 3 Performance using Full Self-Driving (Supervised) versions 12.5.1 and 12.5.3. That worked out to an average of one intervention every 13 miles, according to AMCI—not a universal rate for Tesla FSD or a result independently validated across other vehicles and routes.
What AMCI tested—and what it reported
AMCI Testing said its 2024 evaluation covered more than 1,000 miles (1,610 km) in a 2024 Tesla Model 3 Performance. The car ran FSD software versions 12.5.1 and 12.5.3 across city streets, rural two-lane highways, mountain roads, and freeways. AMCI reported more than 75 interventions, averaging one every 13 miles. AMCI’s September 24, 2024 release described the system as impressive in some circumstances but not consistently reliable.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
I Llike Simulator Safety Driving | $8.87 | Buy on Amazon |
That rate reflects AMCI’s particular vehicle, software, route mix, drivers, and intervention rule. It should not be read as a standardized industry metric or as a like-for-like comparison with another automated driving system.
What AMCI counted as an intervention
In an October 2 follow-up, AMCI said an intervention occurred when a driver took control and forced FSD to disengage because its actions put occupants, the public, or other motorists at risk. This definition is specific to AMCI’s protocol; the releases do not establish that it is a standard metric shared across systems. AMCI’s explanation of its intervention rate provides that operational definition.
#1 Best Overall
Examples shown in the released videos
A Futurism report covering the test described selected video examples: FSD ran a visible red light at night, and on a curve it failed to recognize or follow a double-yellow line, veering toward oncoming traffic before the driver intervened. The report also noted a successful maneuver in which the Tesla moved aside on a narrow road to let oncoming traffic pass.
These are examples from released footage, not a complete event-by-event audit of the drive. AMCI Testing director Guy Mangiamele characterized the system’s inconsistency this way: “What’s most disconcerting and unpredictable is that you may watch FSD successfully negotiate a specific scenario many times — often on the same stretch of road or intersection — only to have it inexplicably fail the next time.” That is Mangiamele’s assessment, not a regulator’s finding.
Why AMCI said the results were concerning
AMCI Global CEO David Stokols said in the September 24, 2024 release: “Although it positively impresses in some circumstances, you simply cannot reliably rely on the accuracy or reasoning behind its responses.” His comment summarizes AMCI’s judgment of this test. It does not establish how FSD performs across all drivers, software versions, vehicles, or roads.
Keep the later New York City test separate
AMCI issued a separate release on October 10, 2024, describing a 300-mile New York City assessment in a Hardware 4-equipped 2024 Model 3 running software version 12.5.4. AMCI reported an average intervention rate greater than once every 11 miles in that test. The NYC release covers a different distance, location, vehicle configuration, and software version; its rate is not part of the earlier 1,000-mile result.
What this test can—and cannot—show
The AMCI result is useful as a report of what one testing organization observed under its protocol. It is not enough by itself to establish a general intervention rate for Tesla FSD or to rank it against a competing system. A meaningful comparison would need to align software and vehicle hardware, geography and road types, distance, route and driver selection, the intervention trigger, and whether the result counts interventions, measures a rate, or records disengagements. AMCI called for agreed metrics, but its releases do not provide a full event-level dataset or a directly matched competing-system evaluation.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
