Digital transformation has been a significant force shaping innovation and R&D processes. Big data, cloud computing, and Industry 4.0, along with leading artificial intelligence applications, have been the driving forces behind this process. In this context, AI agents and multi-agent systems, the next generation of AI managers, have begun to become the invisible but increasingly important actors of innovation. Agents have started to play a significant role in everything from literature reviews and experimental model design to analysis, system training, reporting, and even market research. As a result of these processes, agents have begun to accelerate innovation and R&D activity cycles.

This research first summarizes classical innovation approaches such as linear, chain-linked, systemic, open, and user innovation; these approaches are reinterpreted from the perspective of agent-based artificial intelligence. Then, the conceptual framework of AI agents is outlined, and the basic components of LLM-based autonomous agents and multi-agent systems are discussed. The study then discusses the roles of agents in the R&D cycle in the contexts of idea generation, literature and patent search, design and simulation, testing and validation, commercialization, and market analysis. Human-agent collaboration models, multi-agent orchestration, and ethical and governance dimensions are examined; and example application scenarios from university-industry collaboration, SMEs, healthcare, energy, and manufacturing sectors are presented. Finally, a future-oriented perspective is developed within the framework of the self-driving laboratories approach and agent-based AI-focused corporate transformation trends. This research argues that artificial intelligence agents are not authorities replacing human intelligence, but rather "intelligent partners" that, when properly designed, accelerate innovation.