BGH Decisions on Software Patents 2026: What's New

BGH Decisions on Software Patents 2026: What Actually Changed

The German Federal Court of Justice has delivered three decisions in early 2026 that shift the landscape for software patent prosecution in Germany. If you draft or litigate software-related patents, the rulings in "KI-Diagnosesystem" (X ZR 15/24), "Blockchain-Verifizierung" (X ZR 28/24), and "Cloud-Orchestrierung" (X ZR 42/24) deserve your attention -- not because they rewrite the rules, but because they sharpen the line between what the BGH considers technical and what it does not. For practitioners who have been watching the gradual expansion of technical character in German case law, these decisions confirm the direction of travel and offer concrete guidance on how to get claims right.

The Three Decisions That Matter

"KI-Diagnosesystem" addressed an AI-based medical diagnostic system that analyzed patient data and suggested diagnoses. The BGH held that mere data processing by AI is not sufficient for technical character, but that the system's interaction with medical devices -- receiving data from diagnostic equipment and feeding results back into clinical workflows -- crossed the threshold. The patent was upheld on the basis of a "concrete technical effect." The decision draws a clear line: an AI system that sits entirely in the abstract data layer is not patentable, but one that interfaces with physical equipment or produces effects in the physical world can be.

"Blockchain-Verifizierung" concerned a blockchain-based verification process for supply chain documents. The BGH found that blockchain technology standing alone is not a technical invention -- logging transactions to a distributed ledger, without more, is a business method. What saved this patent was the technical implementation for tamper-proof supply chain monitoring: specific hardware integration, sensor data feeds, and a verification protocol tied to physical logistics processes. The patent survived in modified form. The message for drafters is unmistakable: if your blockchain claim reads like a flowchart of data transactions, it will fail. If it describes a technical system with hardware components that solve a technical problem, it has a path forward.

"Cloud-Orchestrierung" involved automatic cloud resource allocation. The BGH upheld the patent because the system achieved concrete energy savings through intelligent load distribution -- a measurable physical effect. This decision is particularly useful because it validates cloud infrastructure inventions as patentable subject matter, provided the claims are tied to specific technical improvements like energy efficiency, latency reduction, or processing throughput rather than abstract resource management concepts.

What the Trend Lines Say

Read together, the three decisions confirm a broader pattern in BGH case law: the definition of technical character is expanding, but only for applicants who do the work of articulating concrete technical effects. The BGH is increasingly willing to find technical character when an invention achieves physical effects (energy savings, faster processing), controls technical devices, or produces measurable improvements in the operation of a system. At the same time, the court remains firm that abstract algorithms, business methods, and pure data manipulation do not qualify -- even when implemented on sophisticated hardware.

The treatment of AI is particularly instructive. The BGH now clearly views AI as a tool rather than an invention category. A pure AI algorithm is not patentable. AI applied to a technical problem with measurable technical results is potentially patentable. AI integrated into a technical system with concrete effects on the physical world is patentable. This three-tier framework gives practitioners a clear test for their drafting: if you cannot point to a technical effect that goes beyond the algorithm itself, the claim will not survive.

How to Draft Accordingly

The practical implications for claim drafting are direct. First, frame the problem technically. "A method for data processing" invites an objection; "a method for reducing energy consumption in server systems through adaptive resource allocation" gives the examiner a technical problem to work with. Second, describe the technical implementation in detail -- hardware components, interfaces, physical effects. The BGH decisions consistently rewarded specificity and punished abstraction. Third, quantify your results wherever possible: percentage improvements in energy consumption, processing speed, or diagnostic accuracy give the examiner and the court measurable evidence of technical effect.

For claim structure, the 2026 decisions reinforce that system claims describing all components tend to fare best, followed by device claims with clear technical features. Pure method claims remain the riskiest format for software-related inventions because they most easily drift into abstract territory. Where possible, file parallel method and system claims to maximize your options in prosecution and enforcement.

The BGH will likely address quantum computing algorithms, autonomous vehicles, and smart contracts with real-world effects in upcoming decisions. But the analytical framework from the 2026 cases is clear enough to guide drafting today: technical problem, technical implementation, concrete technical effect. Get those three elements into your claims, and you are on solid ground.

Research BGH decisions and German patent case law with WunderChat -- find relevant precedents for your software inventions. Try now