Generative AI boosts digital twins across healthcare, smart cities and supply chains

By integrating multimodal data streams, AI-driven digital twins enhance urban resilience, optimize energy flows, and improve supply chain responsiveness. The result is faster decision-making, more robust planning, and the ability to test solutions virtually before deploying them in real environments.

Generative artificial intelligence is reshaping the future of industrial systems, with far-reaching implications for manufacturing, healthcare, smart cities, and global supply chains. A new study by researchers from Kazimierz Wielki University in Poland maps how generative AI is transforming AI-based digital twins, offering both economic opportunities and serious ethical challenges.

Published in Applied Sciences, the study Role of Generative AI in AI-Based Digital Twins in Industry 5.0 and Evolution to Industry 6.0 provides a bibliometric review of recent research, analyzing contributions from 2023 to 2025. The authors argue that generative AI already strengthens digital twins by producing synthetic data, simulations, and design scenarios, and they forecast that its influence will accelerate the transition from Industry 5.0’s human-centric focus to Industry 6.0’s autonomous, hyperconnected ecosystems.

How generative AI enhances digital twins today

The review identifies generative AI as a powerful enabler of digital twin technology, giving these virtual replicas the ability to adapt and evolve beyond static models. By generating synthetic scenarios, data sets, and design options, generative AI expands the scope of what digital twins can simulate, predict, and optimize.

The authors explain that in Industry 5.0, digital twins are being used to improve sustainability, human-machine collaboration, and efficiency. Generative AI boosts these functions by making twins more adaptive and less dependent on costly real-world prototyping. For example, in industrial cyber-physical systems and the Industrial Internet of Things, generative AI helps predict equipment performance, optimize resource allocation, and simulate factory layouts. In healthcare, it aids patient-specific digital twins for imaging and clinical trials, paving the way for personalized medicine.

The review also shows how generative AI supports smart cities, energy systems, and logistics. By integrating multimodal data streams, AI-driven digital twins enhance urban resilience, optimize energy flows, and improve supply chain responsiveness. The result is faster decision-making, more robust planning, and the ability to test solutions virtually before deploying them in real environments.

What barriers and risks threaten adoption

Further, the study stresses that deploying generative AI in digital twins faces significant barriers. One of the most pressing issues is the cost of computing infrastructure. Training and running generative models requires high-performance hardware and energy resources that many organizations cannot easily afford.

Data quality is another challenge. Generative AI can only enhance digital twins if it is trained on reliable, representative data. Inconsistent, biased, or incomplete datasets risk producing flawed simulations that could undermine trust and lead to costly errors in sectors like healthcare and manufacturing.

The authors also identify technical bottlenecks, including difficulties in integrating generative AI with legacy systems, ensuring scalability, and protecting against cybersecurity threats. The opacity of generative models compounds the problem, as explainability remains limited and regulators demand transparency in high-stakes applications.

Beyond technical barriers, the review highlights societal and ethical risks. Privacy is a recurring concern when digital twins integrate personal or sensitive data, especially in healthcare. Intellectual property disputes loom over content produced by generative models. Environmental sustainability is another issue, given the high energy consumption of training large AI models. Finally, the risk of a widening digital divide threatens equitable access, as wealthier regions may benefit disproportionately from these technologies while others fall behind.

What Industry 6.0 will look like with generative AI

Further, the authors describe how generative AI and digital twins are poised to drive the shift from Industry 5.0 to Industry 6.0. In their vision, Industry 6.0 will be defined by autonomous, self-organizing, and hyperconnected ecosystems where digital twins learn, adapt, and collaborate with minimal human oversight.

The review points to several promising directions. Generative AI could automate the creation and continuous updating of digital twins, reducing the time and cost of model development. Real-time anomaly detection and predictive maintenance will become more reliable, helping industries reduce downtime and prevent costly failures. In supply chains, AI-driven twins could simulate logistics scenarios under variable conditions, improving resilience against disruptions. In education and training, generative AI could make digital twins more interactive, creating immersive environments for learning and skill development.

The authors argue that these advances will bring economic benefits such as reduced prototyping costs, optimized logistics, and productivity gains. But they caution that the same technologies raise governance questions that cannot be ignored. Accountability for autonomous decisions, bias in generative outputs, and the energy footprint of large-scale AI adoption will require international regulation and industry standards.