Software can still be patented, but after Mayo and Alice every claim must pass the two step Alice Mayo test under 35 U.S.C. § 101: is the claim directed to an abstract idea (a bare concept like math or organizing human activity), and if so, does it add an inventive concept (a real technical contribution beyond a generic computer)?
The most important shift for applicants to understand is that examiners are not rejecting software because it is software. They are rejecting claims that describe a business result achieved on a computer, rather than a technical improvement to how the computer itself works. The winning strategy is to frame the invention around a specific technical problem, describe the prior technical approach and why it fell short, then explain the concrete technical mechanism that solves the problem, not just the outcome it produces. Claims should describe the mechanism itself rather than simply claiming that a system is configured to optimize or improve a result. In 2025 and 2026, the United States Patent and Trademark Office issued new internal guidance encouraging examiners to give more weight to technical specificity in the specification, and data shows the appeal reversal rate for these rejections has climbed substantially as a result. For our clients, this means detailed, technically grounded drafting is more valuable than ever, and it is exactly the kind of drafting our software patent practice is built around.
Sources: Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014); Mayo Collaborative Services v. Prometheus Laboratories, 566 U.S. 66 (2012); MPEP § 2106.