Connected cars can provide entertainment, give directions and help prevent collisions.

But according to a new report from AI-powered cloud-based data management platform Upstream, one thing they can do — but aren’t — is help reduce vehicle recalls.

Upstream’s report, Under Pressure: Why After-Sales Quality Strategies Must Evolve in the Age of Connected Vehicle Data and AI, said the causes of 70% of all recalls since 2020 and 89% of those involving electric vehicles could have been detected earlier using connected vehicle signals.

The report is based on analysis of National Highway Traffic Safety Administration data, including 5,189 recall records and more than 30,000 consumer complaints involving model year 2020 vehicles and newer. The study used artificial intelligence tools to identify trends and categorize issues by various factors.

That analysis found the percentage of recalls that could have been detected earlier using connected vehicle signals has steadily increased from 69% in 2020 to 75% so far in 2025. Instead, Upstream said, many manufacturers struggle with “fragmented telemetry, siloed systems and reactive investigation processes” that slow the detection of issues.

“This trend highlights the strategic value of integrating connected data into after-sales quality programs,” the report said, “not just to accelerate investigations but to reduce warranty exposure, limit recall scale and safeguard customer trust.”

The issue is especially pronounced among EVs, whose share of the recall landscape is growing faster than its share of the new vehicle market.

According to the report, EVs account for more than 14% of the recalls issues so far in 2025 thanks to a steady but significant rise from about 2.5% in 2020, which Upstream said indicates “growing exposure to quality issues as EV adoption scales.” In that same span, EV market share increased from 1.6% to about 10%.

Upstream’s analysis showed detectable EV recalls reached or surpassed 90% of all EV recalls in each of the past four years, and 49% of EV-related recalls could have been identified early simply by monitoring the diagnostic trouble code.

“EVs are a clear illustration of both the growing quality risk and the opportunity for early detection,” the report said. “This analysis underscores the critical opportunity for OEMs to integrate connected data analytics into their EV quality programs, transforming reactive recall processes into proactive, data-driven interventions.”

“The automotive industry is navigating a perfect storm of rising warranty costs, growing recall volumes and increasing system complexity, particularly as EVs and software-defined vehicles become the new standard,” Upstream CEO and co-founder Yoav Levy said. “Traditional quality analytics, which rely on retrospective claims and service data, are no longer sufficient. After-sales quality teams need to leverage connected vehicle data and invest in the right AI tools to enable scalable, proactive quality monitoring and detection.

“The opportunity is clear. With the right machine learning foundation and modern data infrastructure, automakers can detect issues earlier, pinpoint root causes faster and ultimately reduce recall scope while protecting budgets and customer satisfaction.”

The full report can be downloaded here.