US Regulators Expand GM V8 Probe After Reported Post-Recall Failures

US safety regulators have expanded their investigation of General Motors’ 6.2-liter L87 V8 from 597,630 to 997,743 vehicles after receiving reports of engine failures following recall work. The broader Engineering Analysis now covers relevant GM trucks and SUVs produced through 2026, including engines built after the period addressed by the original recall.

The central issue is not whether all 997,743 engines are defective; neither the investigation nor GM has established that. The unresolved question is whether the original remedy and production cutoff reliably identified the engines at risk. That distinction matters because reported failures include engines that received the prescribed oil change, replacement engines and engines produced after the recall period.

Why the investigation grew

By August 2026, the National Highway Traffic Safety Administration’s Office of Defects Investigation had received at least 499 complaints alleging failures after recall remedies. Of those, 473 concerned vehicles given a change in recommended oil viscosity, while 26 involved vehicles that had received replacement engines. Regulators also recorded 191 failure reports involving engines produced after the original recall’s production period.

GM separately reported receiving 6,953 complaints of post-remedy engine failure. Complaint totals are inputs to an investigation, not confirmed defect counts, and reports can require further technical validation. NHTSA explains that it assesses complaint frequency, potential severity and other available data through a risk-based defects-investigation process. An Engineering Analysis is intended to examine the scope, consequences and possible remedy of a suspected safety defect in greater depth.

The US population includes L87-equipped Chevrolet Silverado, Tahoe and Suburban models; GMC Sierra 1500, Yukon and Yukon XL models; and Cadillac Escalade and Escalade ESV models from 2021 through 2026. Inclusion in that population means a vehicle falls within the investigation’s scope, not that its engine has been found defective.

Remedy effectiveness is now the systems question

GM attributed the original problem to supplier manufacturing-quality issues involving connecting rods and crankshafts in certain engines. The company emphasized that some L87 engines are affected and others are not. That creates a screening challenge: a remedy must distinguish engines with relevant manufacturing defects from engines that can remain in service without unnecessary replacement.

The original approach depended on dealer inspection. Engines that did not pass were replaced, while those that passed received higher-viscosity oil, a replacement filter and a new oil-fill cap reflecting the revised specification. From an engineering-policy perspective, post-remedy reports test two separate assumptions: whether the inspection correctly classified engines and whether the resulting intervention adequately controlled the failure risk.

The 26 complaints involving replacement engines raise a different scope question. They do not prove that all replacement engines share the original defect, but they give investigators a reason to examine replacement-part production, manufacturing dates and whether the same reported failure mechanism is present. Likewise, the 191 reports involving later-production engines challenge the original date boundary without establishing that every engine built after it is affected.

Production cutoffs are consequential because they determine which owners receive notices, inspections and no-cost recall work. If investigators find that the relevant manufacturing condition continued beyond the original period, owners of later trucks and SUVs could face the same potential loss of propulsion concern without initially being included in the recall. If the later reports have different causes, expanding a remedy without that distinction could impose unnecessary service demand and engine replacements. The investigation must therefore separate a common manufacturing issue from unrelated failures that produce similar symptoms.

The investigation is broader than the recall

GM’s original global recall covered 721,000 vehicles produced from 2021 through 2024. The current US Engineering Analysis reaches nearly one million vehicles, but GM’s US recall population remains 597,630. An investigation can be wider than a recall because regulators need a comparison population to determine where a defect begins and ends.

The issue also extends beyond the United States. Australia recalled the L87 in May 2025, with GM Australia and New Zealand offering inspection and, where necessary, engine replacement for as many as 9,468 Chevrolet Silverado 1500s. Australia’s official notice states that manufacturing defects in connecting-rod or crankshaft components could cause engine damage or failure. GM ANZ had not confirmed local owner complaints or what the expanded US investigation could mean for Australian vehicles.

NHTSA’s next task is to determine whether the reported post-remedy and later-production failures share the manufacturing issue identified in the original recall. Until that link and its boundaries are established, 997,743 is the investigation population not a count of defective engines but the reports following oil changes and complete replacements make remedy effectiveness the decisive question.

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