Artificial Intelligence (AI) is often considered a general purpose technology with the potential for disruption in most existing markets. When facing a substantial technological disruption, it has been shown that incumbents are confronted with many challenges to develop new products (conflicting customer incentives, cognitive lock-in, lack of complementary assets), despite that this activity is critical to ensure companies' survival. Overcoming those issues lies with companies' capacity to strategically renew their resources and remain competitive in their market. However, while the literature has shown how incumbents used AI technologies to increase efficiency or productivity, little is known about how they rely on AI technologies in their inventive process. It has also been observed that companies broadly vary in the ways they are embracing AI technologies, but most studies approach AI-based inventions as homogeneous. Hence, to cope with this research gap, we adopted a strategic renewal perspective (e.g., Agarwal & Helfat, 2009), and we conceptualized a new taxonomy of different types of AI technology usages in companies' inventive processes. Creative AI-based inventions focus on implementing AI technologies that serve new-to-company applications and open up new technological paths. Follow-up creative AI-based inventions indicate the company's efforts in following the newly created technological path. Low AI-based inventions specify the usage of AI to extend the existing technology portfolio without disrupting the current business of the company. By drawing upon an example from the agricultural sector, we developed techniques relying on patent data to implement our taxonomy by dealing with 39,000 utility patents from the sector's top-19 companies. First, we explored how incumbents and new entrants from different industry origins relied on different AI technology usages with different adoption patterns. Second, by relying on patent indicators, we reviewed the performances of the different types of AI technology usages in companies' inventive process. We contribute to strategic renewal theory by revealing that incumbents from the industry we focus on others rely on AI to renew their patent portfolio to strategically develop new business. Second, we demonstrated the co-existence of different usages of AI technologies in companies' inventive processes to complement incumbents' portfolios and open new innovative paths. Consequently, companies should plan which type of strategic renewal is appropriate for their organisations based on their strategic playing field to achieve the desired outcome.