Infosys Manufacturing Tech Index: Why AI has moved from pilot to production imperative

Artificial intelligence is no longer a future bet for manufacturers, it is rapidly becoming foundational to how factories operate, secure themselves, and compete. According to the inaugural Infosys Manufacturing Tech Index: AI Pulse, 75% of global manufacturers have now embedded AI into their enterprise strategy, signalling a decisive shift from experimentation to execution.

Based on insights from 650 senior manufacturing executives across regions and industries, the study captures a sector at an inflection point, one where ambition is high, investments are rising sharply, but the path to value remains uneven and fraught with execution challenges

From experimentation to enterprise strategy

The report highlights that manufacturers are no longer debating whether to adopt AI, but how fast and how deeply it can be integrated into core operations. With rising cost pressures, labour shortages, and increasing operational complexity, traditional automation and human decision-making are proving insufficient. AI is now being positioned as a critical lever to reduce unit costs, accelerate innovation cycles, and respond dynamically to market and supply-chain volatility.

However, the research also reveals a crucial nuance: while strategic intent around AI is strong, strategy alone does not guarantee success. Manufacturers that declare AI a critical pillar tend to launch significantly more initiatives, but their success rates are not materially higher than peers. This suggests that execution capability, not intent, is the real differentiator.

Big investments, bigger expectations

Unlike software-led industries, AI in manufacturing is capital-intensive. More than half of manufacturers surveyed are investing over $2 million per AI initiative, with a median spend ranging between $2 million and $2.5 million. These costs extend far beyond model development, encompassing data engineering, system integration with physical assets, cybersecurity safeguards, and workforce enablement.

The scale of investment underscores a clear message: AI in manufacturing is no longer a lightweight digital experiment. It demands the same governance discipline, value-tracking rigor, and capital-allocation scrutiny as any other core operational program. Yet, despite this financial commitment, only about one in five AI initiatives have begun meeting business objectives, a gap that continues to challenge leadership confidence.

Cybersecurity: both the front-runner and the roadblock

Cybersecurity has emerged as the most common AI use case in manufacturing, with 57% of respondents deploying AI across cybersecurity and operational technology (OT) systems. As IT and OT environments increasingly converge, AI is being used to monitor vulnerabilities, detect anomalies, and prevent disruptions before they escalate into plant-wide failures.

Ironically, cybersecurity is also cited as the single biggest barrier to scaling AI, flagged by 23% of respondents. As AI systems gain more autonomy and become embedded into physical operations, the risk surface expands dramatically. Without unified visibility, strong governance, and secure IT-OT integration, manufacturers find it difficult to scale AI safely across the enterprise. Data challenges, ranging from quality and access to lineage and governance, closely follow as the second-largest inhibitor.

A venture capital–style AI portfolio

One of the more striking insights from the report is how AI initiatives behave like a venture capital portfolio. While a small percentage of deployments generate meaningful value, many fail to deliver returns, and a significant number are cancelled either before or after deployment. Roughly 60% of initiatives are still stuck in planning, proof-of-concept, or pilot stages.

This uneven value realisation is not entirely unexpected. AI in manufacturing often operates deep within production systems, making outcomes highly context-specific. The research suggests that organisations must become more disciplined, using tighter stage gates, stronger operating models, and the courage to shut down underperforming initiatives and reinvest in those showing promise.

The great divide in AI sentiment

Despite rising investment and strategic focus, sentiment around AI remains sharply polarised. Nearly one-third of manufacturing leaders believe AI’s value is transformational, while an almost equal proportion feel its impact is heavily overstated. This divide reflects varying levels of maturity, execution readiness, and prior success, or failure, with AI programs.

Interestingly, manufacturers that embed AI into their enterprise strategy display greater confidence in AI’s long-term value, even though skepticism persists within this group. The report warns that unchecked skepticism can stall adoption, while organisations that remain stuck in experimentation risk falling behind competitors who are compounding their AI advantage through early, sustained investment.

What separates leaders from laggards

The central takeaway from the Infosys Manufacturing Tech Index is clear: the next phase of AI maturity will be defined not by ambition, but by execution. Manufacturers that see tangible returns are those pairing AI strategy with unified data architectures, robust cybersecurity frameworks, cross-functional operating models, and systematic workforce readiness programs.

As AI moves deeper into shop floors, supply chains, and customer-facing processes, success will hinge on how effectively organisations integrate AI into everyday decision flows, not how many pilots they launch. For manufacturers, the race is no longer about adopting AI first, but about operationalising it better than the rest.