AI in manufacturing set to unleash new era of profit
Subscribe to our free newsletter today to keep up to date with the latest manufacturing news.
A growing number of manufacturers are placing bold bets on artificial intelligence. According to the Future-Ready Manufacturing Study 2025 by Tata Consultancy Services and AWS, manufacturers now expect AI to deliver measurable improvements to profitability in as little as two years.
The study reveals that 88 percent of industry leaders believe AI will add at least 5 percent to their operating margins, while one in four expects returns exceeding 10 percent. These expectations have triggered a major reallocation of capital. More than half of all transformation spending is now directed toward AI and autonomous systems, far outpacing investments in workforce reskilling or cloud infrastructure.
This strategic shift signals the arrival of a new manufacturing playbook, one in which predictive algorithms, autonomous agents and intelligent infrastructure become the backbone of performance. Yet despite the ambition, challenges around data readiness, system integration and operational trust suggest the industry is not fully prepared to reap the rewards.
Investment is up, but infrastructure lags behind
Executives are clear about what they want. Seventy-five percent of survey respondents said AI would become one of the top three contributors to their bottom line by 2026. By then, they expect intelligent systems to manage key workflows, forecast disruptions and optimize performance without constant human oversight.
The momentum is unmistakable, but the foundation remains fragile. Only 21 percent of manufacturers report being fully AI-ready, defined as having unified, clean and contextual data systems across all plants. The remaining majority operate with partial or fragmented datasets, with inconsistent quality, siloed structures and legacy systems blocking full integration.
This lack of foundational readiness creates friction across the enterprise. Many advanced models require broad access to real-time inputs from supply chains, production equipment and quality control systems. In the absence of these connections, AI cannot deliver the automation and agility it promises.
Technical debt is also a major concern. More than half of respondents cite integration with legacy systems as their top obstacle. In many plants, decades of software layering have created brittle environments where modern AI agents cannot function at scale. Security concerns compound the challenge, especially in sectors where a cyber breach could disrupt physical operations or endanger workers. Fifty-two percent of executives identified plant-level security and governance as key hurdles to deployment.
AI promises smarter factories, but trust remains an issue
Even among early adopters, there is a noticeable gap between what is implemented and how decisions are made on the factory floor. Despite heavy investments in digital infrastructure, operational behavior often reverts to traditional, analog hedges when volatility strikes.
Following recent supply chain disruptions, 61 percent of manufacturers increased safety stock levels. Half turned to multi-sourcing strategies to mitigate risk. In contrast, only 26 percent reported using scenario planning or digital twin simulations to guide response. This reluctance to rely on AI-generated insights shows that manufacturers still trust physical buffers over digital forecasts.
Ozgur Tohumcu, general manager for automotive and manufacturing at AWS, describes the moment as a turning point. “Manufacturers today are facing unprecedented pressure — from tight margins to volatile supply chains and workforce gaps. By embedding artificial intelligence into every layer of the operation and leveraging cloud-native architecture, manufacturers can move beyond simple automation to true autonomous decision-making.”
Yet trust cannot be automated. To build confidence in intelligent systems, manufacturers must demonstrate reliability through staged deployment. For now, AI agents are mostly assigned to administrative tasks. Two-thirds of companies already use or plan to use agents for routine work order approvals, and more than 70 percent expect agents to handle half of production decisions by 2028.
Agentic systems and a hybrid future
The industry is gradually shifting from AI as a tool to AI as a collaborator. This evolution, known as agentic AI, refers to systems capable of taking independent action with minimal supervision. Unlike traditional analytics, agentic models can interpret real-time data, apply rules and execute decisions dynamically. In early deployments, such systems have improved productivity in knowledge-intensive roles such as quality inspection and IT support.
However, adoption remains uneven. While 49 percent of firms report efficiency gains among quality inspectors, only 29 percent say the same for frontline roles like maintenance technicians. As with many digital transformations, cognitive tasks tend to automate before those requiring physical coordination.
Platform strategy is another key decision. Most manufacturers are steering away from single-vendor ecosystems. Sixty-three percent prefer hybrid or multi-platform approaches, ensuring flexibility and leverage. A third plan to coordinate through multiple platform-native agents, while another 30 percent will blend custom and commercial orchestration tools.
Turning ambition into returns
If manufacturers want to turn their AI investments into lasting profit, they must address foundational gaps. The first priority is data modernization. Without high-quality, integrated data, even the most advanced models cannot function effectively. Second, companies must invest in building trust across their operations. This includes demonstrating the effectiveness of AI in low-risk areas before extending control to more complex functions.
Finally, strategic diversity in platform architecture will help firms remain agile. Locking into a single vendor can slow innovation and reduce bargaining power.
AI is no longer a side project in manufacturing. It is a core strategy. However, converting that strategy into profit will require patience, investment and a focus on the basics, especially data readiness, cross-functional coordination and operational trust.