Why Automotive Leaders Are Trading AI Hype for Agentic Reality

If 2025 was the year of AI experimentation, 2026 is shaping up to be the year of the Agentic shift. In a recent webinar hosted by KPMG and SAP titled “The Promise for AI in Automotive Companies,” industry veterans stripped away the marketing gloss to show what is actually happening on the factory floor and in the boardroom of automotive companies. The consensus? Unregulated AI pilots are becoming a thing of the past. In its place, a more disciplined, high-stakes reality is emerging where AI is an active agent in the supply chain.

The Rise of AI Agents

The adoption of Agentic AI has been rapid. According to KPMG’s recent polling data shared during the session, the industry has aggressively pivoted from tentative pilots to scalable deployment in less than six months.

Hernan De la Torre, Principal at KPMG, highlighted this acceleration, noting that while less than 15% of companies were deploying AI at scale earlier in the year, that number has surged. “What we found is in just a short amount of time, up to 42% of the companies we survey are now leveraging agents,” De la Torre revealed. “Organizations are no longer looking at the question ‘should we,’ but rather, ‘how fast can we scale?’” ​

This change concerns systems capable of autonomous decision-making in complex environments, such as navigating the volatile new normal of global tariffs. Instead of relying on static spreadsheets, automotive leaders are using these agents for real-time scenario simulation, allowing them to dynamically adjust sourcing strategies before a new policy even goes into effect.

Specialization Over Displacement

Still, the conversation repeatedly returned to a critical, often overlooked truth: AI is a liability without human governance. The fear of mass displacement is being replaced by a more nuanced reality of role evolution. The crush in the middle of management may feel the pressure to adapt, but the need for specialized human insight has never been higher.

“It’s not going to be the machines delivering autonomous vehicles,” De la Torre emphasized. “It will be very smart people putting their thoughts around how they are going to operate.”

This sentiment was echoed by SAP’s experts, who warned against a band-aid approach of slapping AI onto broken processes without re-engineering workflows or upskilling the workforce.

Strategic Capability vs. Shiny Tools

The discussion indicated that CIOs and supply chain leaders in the automotive sector should stop chasing use cases that lack a balance sheet impact. The most successful implementations target high-impact areas, such as risk analysis in tier-2 supplier networks, rather than low-value administrative tasks.

Andrea Haupfear, AI Business Process Architect at SAP, summarized this shift perfectly. “The future of AI isn’t just about those smarter algorithms, it’s about those smarter organizations that ultimately will embrace AI not just as a tool, but as a strategic capability,” she said. “Those will be the ones that will ultimately be shaping the next wave.”

Thus, as we move deeper into 2026, the differentiator will no longer be who has AI, but who has the governance to trust it. To this end, the experts from KPMG and SAP advised organizations to:

• Audit their pilots: If a project doesn’t have a direct line to P&L impact, pause it.
• Focus on data: An agent is only as good as the data it is fed. If master data is messy, AI will simply make bad decisions faster.
• Empower the human loop: Invest in the smart people who will audit the machines.

What This Means for SAPinsiders

Distinguish between embedded efficiency and domain-adapted advantage. As Bill Newman, Industry Executive Advisor for SAP noted, while SAP provides hundreds of embedded AI capabilities for routine tasks like auto-reconciliations, the real competitive edge lies in domain-adapted models. Additionally, the high-value work involves configuring tailored models to address complex automotive challenges, such as the Tier 1-to-Tier 2 supplier pricing optimizer mentioned by Haupfear. This tool uses agents to recommend optimal sourcing based on tariffs, lead times, and on-time in-full metrics, offering a genuine secret sauce for competitive differentiation.

Stop automating waste. Before spinning up an SAP Business Technology Platform (BTP) AI service, organizations should conduct a ruthless process audit. As Rich Wilkins of SAP advised, adopt a dissect-and-enable approach by breaking a primary process into sub-processes. Next, identify where the actual waste sits, and only then apply AI to remove it.

Prepare data for agentic action. SAPinsiders must pivot their data strategy from reporting readiness to execution readiness. An analytic dashboard can tolerate a small margin of error; an autonomous agent booking a supplier order cannot. Therefore, data foundation and governance protocols must be robust enough to support human-in-the-loop verification rather than just historical analysis.

Some or all of the services described herein may not be permissible for KPMG audit clients and their affiliates or related entities.