GDM Holding filed 2 patents in H1 2026, all in AI & Machine Learning.
The AI & Machine Learning filings cover gaming applications of machine learning, including a multi-agent reinforcement learning system that trains AI policies by dynamically pairing agents with diverse opponents for game AI and simulation purposes. The category also includes a companion agent that processes gameplay video streams to answer player questions and overlay visual hints in real-time.
The AI & Machine Learning category accounts for 2 patents, both centered on applying machine learning to interactive game environments. The first describes a reinforcement learning system where each agent in a training pool maintains its own distinct distribution over potential opponents, rather than a single shared matchmaking policy. This per-learner approach, combined with varied hyperparameters and the seeding of fixed-policy opponents, allows the collective pool to improve over time without converging on one dominant strategy. The second patent covers a companion agent that monitors a gameplay video stream from outside the game itself, requiring no integration from the developer, and uses session memory, interaction rules, and tool-use within a unified pipeline to answer player questions and place visual hints directly onto the screen in real-time.
All data sourced from USPTO patent filings. Google Patents may take several weeks to index recent publications. If a link is unavailable, search for the patent number at USPTO Patent Public Search.