From Price Takers to Price Makers? The Welfare Implications of AI-Driven Batteries in Electricity Markets

High electricity prices and the integration of intermittent renewables have made large-scale battery storage a key technology for the energy transition, particularly in Austria and Germany. Yet the welfare impact of batteries hinges on the strategic behaviour of their operators. This project asks whether profit-maximising, AI-driven batteries act as competitive price takers, enhancing market efficiency, or learn to exercise market power as strategic price makers, eroding social gains. First, the analysis establishes a benchmark for the maximum changes in producer and consumer surplus from storage under socially optimal operation, simulating the efficient intraday reallocation of electricity for different battery capacities. The key methodological contribution is the explicit modelling of complex bidding formats, such as combinatorial block bids, which are prevalent in European day-ahead markets. These bids are crucial for understanding how batteries affect prices and welfare by displacing conventional technologies, such as thermal plants or pumped hydro, that typically bid in blocks. Artificial intelligence is then used to simulate the bidding behaviour of profit-maximising battery operators in day-ahead auctions. This enables the identification of market conditions under which operators shift from price-taking to strategic price-making, as well as assessing the effects on prices and market efficiency. Finally, these insights are translated into policy design by assessing the optimal design of network charges to steer dispatch and proposing a centralised auction for grid connections to enable a fast, efficient, and competitive rollout of storage. The aim is a regulatory framework that incentivises batteries to contribute to an efficient, affordable, and sustainable energy system.