NHTSA Expands GM Engine Probe to Nearly One Million Vehicles

Federal safety regulators have expanded their examination of General Motors’ 6.2-liter V8 to nearly one million trucks and SUVs after owners reported engine failures despite completing an earlier recall procedure. The August 21 engineering analysis covers an estimated 997,743 model-year 2021–2026 Chevrolet, GMC and Cadillac vehicles, including engines produced beyond the original recall window.

The immediate safety concern is a reported “loss of motive power due to engine failure.” The investigation file lists at least one crash and one injury associated with sudden power loss. Owners have also alleged severe engine knock and complete shutdown while driving, although the National Highway Traffic Safety Administration has not determined that every complaint has the same cause or that GM’s existing remedy is inadequate.

Why the completed recall remains under examination

GM recalled 597,630 U.S. vehicles from model years 2021–2024 in April 2025. The affected nameplates were the Chevrolet Silverado 1500, Tahoe and Suburban; GMC Sierra 1500, Yukon and Yukon XL; and Cadillac Escalade and Escalade ESV.

The dealer procedure divided engines into two groups. Dealers checked for diagnostic trouble code P0016, which can identify a timing relationship outside the expected range. Engines that failed the inspection were eligible for replacement. Vehicles that passed received 0W-40 oil instead of 0W-20, along with a replacement oil-filler cap and revised owner information.

That was not simply a routine oil recommendation. In engineering terms, moving to a higher-viscosity lubricant changes the oil film available to separate loaded rotating surfaces. It can provide additional protection under some operating conditions, but it does not reverse a crankshaft dimensional error, correct an irregular journal surface or remove all contamination already present inside an engine. Whether the lubricant change adequately controls the documented risk across the treated population is therefore a central question, not a settled conclusion.

The complaint distribution explains why regulators are looking beyond whether dealers followed the written procedure. NHTSA has received 499 allegations of failure after recall work: 473 involved vehicles given the oil-viscosity change, while 26 involved replacement engines. GM’s customer records contain another 6,953 post-remedy complaints. Complaint totals are investigative inputs rather than confirmed defect counts, but their volume gives regulators a basis to test the remedy against field experience.

The production boundary is also in question

The original recall connected the failures to supplier manufacturing and quality problems, including crankshafts with out-of-specification dimensions or surface finishes, rod-bearing damage, and machining sediment in connecting rods and crankshaft oil galleries. GM continues to attribute the problem to those manufacturing defects rather than a fundamental flaw in the L87 engine design.

That distinction matters because a manufacturing defect should ordinarily be bounded by affected parts, suppliers, processes and production dates. NHTSA has received 191 complaints involving engines built after May 2024, outside the original production window. The expanded analysis can examine whether the initial boundary accurately captured the affected output, whether manufacturing controls prevented recurrence and whether later engines share the same failure mechanism.

It can also separate several questions that complaint counts alone cannot answer. Investigators need to determine whether failed engines had the identified crankshaft or bearing conditions, whether the diagnostic screen reliably distinguished engines requiring replacement, and whether replacement-engine complaints arose from the same defects. Without that component-level work, post-remedy failures establish a reason for scrutiny but do not establish one universal cause.

An engineering analysis is not another recall

NHTSA describes an engineering analysis as an investigation stage that may involve additional manufacturer information, inspections, testing or other technical work. The agency says it generally seeks to complete this type of analysis within 18 months, although that is a process target rather than a promised decision date. Its public investigation and recall resources explain that an analysis can close without further action or support a recall if the agency identifies a safety-related defect presenting an unreasonable risk.

No expanded recall has been ordered, and GM has not enlarged the existing campaign. The company says it is cooperating with the evaluation. For engines that passed the original inspection and later fail, GM has established special coverage lasting 10 years or 150,000 miles from the vehicle’s original in-service date.

The decisive issue is now broader than whether thicker oil can add bearing protection. Regulators must determine whether the original diagnostic screen, lubricant change, replacement-engine process and production cutoff collectively addressed the manufacturing-related failure population. Until that work is complete, the existing recall remains in force but its technical scope and remedy are no longer the final word.

By Thomas Caldwell — AMI’s senior editor for mechanical and mobility engineering, covering vehicle electronics, systems integration, electrification, chassis systems, propulsion, and safety policy.