ChatGPT era marks turning point for AI-driven innovation management

A new review published in Sci finds that generative AI has pushed innovation management into a new phase, one in which AI systems are increasingly positioned as collaborators in strategy, product development, sustainability, organizational change, entrepreneurship, and research.

The study, titled How Artificial Intelligence Is Reshaping Innovation Management: Evidence from Pre- and Post-Generative AI Research, analyzed Scopus-indexed studies to compare innovation management research before and after the public rise of ChatGPT, identifying a marked shift from analytics, automation, and decision support toward generative AI-enabled creativity, strategic redesign, and human-AI collaboration.

The authors reviewed 48 studies after screening the literature for relevance and quality. The analysis retained 33 post-ChatGPT studies from 2023 to 2025 and 15 pre-ChatGPT studies from 2020 to 2022. The post-ChatGPT period showed a sharp rise in research output, with the number of reviewed studies growing strongly by 2025. The growth reflects the speed at which businesses, researchers, and policymakers have begun reassessing the role of artificial intelligence in innovation.

AI moves from automation to business model reinvention

The review identifies six major dimensions of AI-enabled innovation management: strategic and business model innovation, product and service innovation, sustainability-oriented innovation, organizational agility and capabilities, human-centric innovation, and knowledge, learning, and research. Across these areas, the authors find a clear transition in how AI is understood and applied.

Strategic and business model innovation emerged as the most represented post-ChatGPT theme. The review finds that AI is increasingly being used to reshape business models, support open innovation, strengthen small and medium-sized enterprises, and help companies operate in uncertain environments. In the pre-ChatGPT literature, AI was more often linked to decision support and strategic foresight. In the post-ChatGPT literature, it is more often connected to active business model redesign.

This change is especially visible in studies on SMEs, energy utilities, real estate consulting, circular business models, and open innovation. AI-based self-assessment tools can help smaller firms evaluate innovation capacity across strategy, alliances, and portfolio management. In industrial settings, AI can support circular business model innovation by helping firms connect digital services, resource efficiency, and sustainability goals. In entrepreneurial settings, GPT-based systems can assist with new service concepts and operational planning.

The review also shows that AI is becoming part of the strategic decision-making process itself. Firms are using AI to analyze complex environments, identify emerging opportunities, refine business choices, and support dynamic capabilities. The authors frame this as a move from AI as a back-office analytical support system to AI as a front-line participant in business transformation.

However, the shift is not without imbalance. The study finds that business model and strategic innovation receive more academic attention than sustainability and human-centered innovation. This suggests that companies and researchers may be moving faster on growth and efficiency applications than on deeper social, environmental, and ethical implications.

In product and service innovation, AI’s contribution is significant but uneven. The review finds that AI is widely applied in the development stage of innovation, where it can assist with prototyping, customer insight analysis, service improvement, forecasting, and concept refinement. Its role in ideation and commercialization is growing, but the evidence remains mixed.

AI can support idea generation and screening, but it does not consistently produce the best ideas or reliably select the strongest ones. That limits claims that generative AI can fully replace human creativity in early-stage innovation. Instead, the evidence points to a hybrid model, where AI accelerates and expands creative work while humans provide judgment, context, and validation.

This is crucial for firms adopting generative AI in product development. Businesses may gain speed and breadth from AI-assisted ideation, but they still need human oversight to assess originality, market fit, feasibility, and ethical risk. The review indicates that generative AI is most useful when embedded into structured innovation workflows rather than treated as an independent creative authority.

Sustainability gains face energy and governance trade-offs

The authors find that AI can support energy-efficient processes, carbon-reduction strategies, renewable energy integration, supply chain innovation, sustainable resource management, and progress toward broader sustainability goals. In SMEs, AI may help firms pursue carbon-neutral business models by improving operations and enabling greener decision-making.

The post-ChatGPT literature gives more direct attention to sustainability than earlier studies, but the field remains underexplored compared with business model innovation and organizational agility. The review notes that pre-ChatGPT studies often treated sustainability indirectly, through efficiency, cost reduction, and uncertainty management. Later studies address sustainability more explicitly, linking AI to emissions management, green innovation, and environmental strategy.

The review simultaneously flags a major contradiction. GenAI may help companies reduce emissions through better supply chain optimization and resource management, but training and operating large AI models can also carry significant energy costs. This creates a tension between AI-enabled sustainability claims and the environmental burden of the technology itself.

Firms cannot assume that AI is automatically sustainable because it improves efficiency in one part of the business. The review points to the need for stronger life-cycle assessment, energy-efficient AI architectures, responsible regulation, and transparent measurement of environmental impact.

Organizational agility is another major area of change. Before ChatGPT, research focused heavily on AI’s ability to improve information processing, reduce labor-intensive work, automate service functions, strengthen forecasting, and support managers across the innovation process. After ChatGPT, the emphasis shifted toward AI as an enabler of collaboration, decentralization, and adaptive decision-making.

The review finds that AI can help organizations become more agile by improving operational efficiency, supply chain management, business intelligence, customer engagement, auditing, tourism services, inventory management, and management practices. Large language models can make business data more accessible by allowing employees to query databases through natural language. AI can also support faster decision cycles by generating analysis, identifying patterns, and reducing manual workload.

However, organizational gains come with risks. The review highlights concerns about overreliance on unverified outputs, ethical blind spots, governance gaps, and uneven managerial readiness. If employees or executives treat AI outputs as authoritative without review, organizations may amplify errors or make decisions based on incomplete assumptions. For this reason, the authors indicate that AI governance must develop alongside technical adoption.

The study also shows that AI is changing the structure of innovation work. By automating routine or repetitive tasks, it can free employees to focus on higher-value creative and strategic activities. However, this benefit depends on organizational design. Companies must decide how to reassign work, train staff, manage accountability, and ensure that AI tools strengthen rather than weaken human judgment.

Human-AI collaboration becomes the next innovation frontier

The review identifies human-centric innovation as one of the most important emerging themes in post-ChatGPT research - an area that examines how AI affects creativity, entrepreneurship, identity, collaboration, and social value creation. The authors distinguish human-AI collaboration from human-centric innovation. Human-AI collaboration refers to operational interaction between people and AI systems during innovation tasks. Human-centric innovation is broader, focusing on how AI can be aligned with human creativity, ethics, and social value.

The literature reviewed by the authors suggests that generative AI is reshaping entrepreneurship by helping users generate data-driven insights, anticipate trends, customize customer interactions, and rethink their own creative roles. Entrepreneurs are increasingly using AI not only as a productivity tool but also as a partner in developing ideas and refining business concepts.

Design thinking is another key area. When combined with AI, design thinking can support more responsive, collaborative, and human-centered innovation processes. In emerging economies, AI-enabled design thinking may help address social needs, bridge socio-economic gaps, and create grassroots entrepreneurial opportunities.

The review points to a broader shift in the theory of innovation. AI is increasingly described as a co-creator rather than a passive tool. That does not mean AI replaces human innovators. Instead, it changes the distribution of work. AI may handle information retrieval, pattern recognition, draft generation, and early concept exploration, while humans focus on judgment, meaning, ethics, context, and strategic direction.

In terms of knowledge and research, AI is becoming both an object of study and a research instrument. It is used to map trends, support strategic foresight, analyze literature, and help researchers examine the relationship between AI, knowledge, and innovation. Earlier research emphasized AI’s ability to reduce uncertainty and detect emerging technologies. More recent work examines how generative AI can support theoretical development, benchmarking, and research synthesis.

The review warns that methodological standards have not kept pace with AI adoption. AI-assisted research tools need transparency, bias controls, and validation. Without these safeguards, generative AI could distort rather than improve knowledge production.

The review does have some limits. The literature is still fragmented, and many studies emphasize positive outcomes while paying less attention to failures, unintended effects, bias, overreliance, and the possible homogenization of creative processes. The study’s own scope is limited by its reliance on Scopus-indexed and open-access publications, which may exclude relevant work from other databases or subscription journals.

In the long run, scholars need more empirical and mixed-method studies on how AI changes innovation teams over time, how human-AI collaboration affects creativity and decision-making, how governance should manage ethical and operational risks, and how different industries and regions adopt AI under varying regulatory and infrastructure conditions.