Manufacturing executives are now directing nearly half of their modernisation budgets toward AI, confident that these systems will deliver profit gains within the next two years.
This decisive shift signals a clear strategic pivot: AI is no longer experimental it is viewed as the core driver of future financial performance.
According to the Future-Ready Manufacturing Study 2025 by Tata Consultancy Services (TCS) and AWS, 88 percent of manufacturers expect AI to contribute at least five percent to operating margins, while one in four anticipate returns exceeding 10 percent.
The capital is available, and the ambition is unmistakable but the infrastructure hasn’t caught up.
A gap persists between bold financial projections and the practical realities on the factory floor. Even as investments in intelligent systems surge, many manufacturers are still held back by fragile data foundations and risk-management processes that depend heavily on costly manual safeguards.
Rising Pressure to Deliver Real AI Value in Manufacturing
The urgency to generate tangible financial returns from technology investments has reached unprecedented levels. According to the study, 75 percent of respondents expect AI to become one of the top three drivers of operating margins by 2026. In response, organisations are directing 51 percent of their transformation budgets toward AI and autonomous systems over the next two years, reflecting a clear shift toward outcome-driven adoption.
This surge in AI investment is overshadowing other critical priorities. Budget allocations for AI now far exceed spending on workforce reskilling (19%) and cloud infrastructure modernisation (16%). For CIOs, the imbalance is a warning sign organisations are trying to deploy advanced algorithms on fragile, outdated systems, creating a structural risk that could undermine the promised returns of AI.
Anupam Singhal, President of Manufacturing at TCS, noted that the sector has long been built on precision, reliability, and a constant drive for higher performance. With the introduction of AI, he said, these strengths are amplified further, enabling more coordinated decision-making and unlocking transformational outcomes rooted in improved predictability, stability, and control.
At TCS, we view this moment as a pivotal opportunity to help manufacturers create resilient, adaptive, and future-ready enterprise ecosystems ones capable of thriving in an era defined by intelligent autonomy.
Analogue Safeguards in a Supposedly Digital Era
Despite substantial investment in predictive technologies, day-to-day operations still reveal a deep trust deficit. When disruptions occur, manufacturers rarely rely on the promised agility of their digital systems. Instead, they fall back on analogue, physical safeguards a clear signal that digital confidence has not yet caught up with digital spending.
After recent disruptions, 61 percent of organisations responded by boosting their safety stock, while half adopted multisourcing strategies for logistics. Yet only 26 percent relied on scenario planning through digital twins to manage volatility underscoring the limited use of the very advanced tools companies are investing in.
This is the core disconnect. Even though 49 percent of respondents cite dynamic inventory optimisation as a key advantage of AI, the dominant instinct remains to stockpile inventory. Supply chain leaders are effectively buying Ferraris but operating them like tractors. Closing this gap will require shifting from reactive, manual safety measures to proactive, system-driven responses.
Ozgur Tohumcu, General Manager of Automotive and Manufacturing at AWS, noted that manufacturers are operating under intense pressure from shrinking margins to unstable supply chains and persistent talent shortages. He said AWS is helping transform the sector through AI-driven autonomous operations, enabling a shift away from manual, reactive processes toward intelligent, self-optimising systems that can operate and scale with far greater efficiency.
By integrating artificial intelligence across every layer of operations and harnessing cloud-native architectures, manufacturers can move beyond basic automation toward true autonomous decision-making where systems can predict, adapt, and act with minimal human input. This shift not only accelerates response times but fundamentally reshapes operations through AI-driven predictability, resilience, and agility.
Infrastructure Debt Holds Back AI Returns
The main barrier to realising AI-driven financial gains isn’t the technology itself it’s the underlying data. Only 21 percent of manufacturers consider themselves “fully AI-ready,” with clean, contextual, and unified data capable of supporting advanced AI models.
The majority of manufacturers (61%) operate with only partial AI readiness, facing inconsistent data quality across different plants. This fragmentation leads to data silos, limiting algorithms from accessing the comprehensive, enterprise-wide information needed for accurate and effective decision-making.
Integration with legacy systems remains the biggest challenge, reported by 54 percent of respondents. This “technical debt,” built up over decades of digitisation, complicates the deployment of modern autonomous agents on older operational technologies.
Security challenges are another major constraint. Security and governance concerns rank as the top plant-level obstacles for 52 percent of respondents. In settings where a cyber-physical breach could halt production or pose physical risks, manufacturers remain cautious about fully embracing autonomous interventions.
The Move Toward Agentic AI in Manufacturing
Despite these challenges, the manufacturing sector is steadily advancing toward agentic AI systems capable of making decisions with minimal human oversight. This shift signals a transformative approach, where intelligent machines begin to autonomously optimise operations and drive business outcomes.
Seventy-four percent of manufacturers anticipate that AI agents will handle up to half of routine production decisions by 2028. In the near term, 66 percent of organisations already permit or plan to permit within the next 12 months AI agents to approve routine work orders without requiring human sign-off.
The evolution from “copilots” to fully autonomous agents capable of completing entire tasks is set to reshape the workforce. While 89 percent of manufacturers anticipate that AI-guided robotics will have an impact, the emphasis remains on augmenting human roles rather than replacing them.
Productivity improvements are currently most pronounced in knowledge-intensive roles. Quality inspectors (49%) and IT support staff (44%) are experiencing the fastest gains, while traditional production positions, such as maintenance technicians (29%), are lagging. This trend reflects a pattern of prioritising cognitive augmentation before extending AI benefits to tasks requiring physical coordination.
As AI agents become integrated across enterprise platforms, architects must decide how best to orchestrate their deployment. The market demonstrates a clear preference for solutions that avoid vendor lock-in.
Sixty-three percent of manufacturers favour hybrid or multi-platform strategies rather than relying on a single vendor. Specifically, 33 percent intend to coordinate operations through multiple platform-native agents, while 30 percent prefer a hybrid approach that combines platform-native tools with custom orchestration. Only 13 percent are prepared to base their operations on a single foundational platform.
Maximising Profit from AI Investments in Manufacturing
To translate this significant AI investment into real profit, the C-suite must move beyond the hype and focus on strategic execution.
The first priority is to fix the data. With just 21 percent of firms fully AI-ready, organisations must focus on modernising their data infrastructure before developing advanced algorithms. Without clean, unified data, high-value applications such as sustainability initiatives and predictive maintenance will struggle to scale effectively.
Second, leaders need to close the AI trust gap. The continued reliance on safety stock reflects a lack of confidence in digital signals. A staged approach to autonomy offers a solution — beginning with administrative tasks such as work orders, where 66 percent of organisations are already moving, before gradually entrusting AI with more complex supply chain decisions.
Finally, avoid the monolithic trap. The data favours a multi-platform strategy to preserve flexibility and control. While manufacturers are investing heavily in AI, achieving real returns depends less on the sophistication of the models and more on the essential groundwork: cleaning data, integrating legacy systems, and cultivating workforce trust.
