AI Patent Insights
Practical guidance from U.S. patent attorneys on patenting AI, machine learning, and software inventions
From Provisional to Non-Provisional: Preserving Support and Priority in the U.S.
Trim the claims if you must, but never the disclosure. Deadlines, claim-by-claim priority, incorporation-by-reference limits, and how to carry provisional claim combinations forward.
Why Provisional Patent Applications Should Use Multiple (and Multiple-Multiple) Dependent Claims
A provisional only protects the combinations it actually supports. A multiple-dependent claim tree is the cheapest way to support them all, especially for AI inventions.
Akamai, AI, and the Three Layers of Liability: Why Your First Drafting Decision Must Be Who You Want to Sue
Divided infringement punishes AI method claims that span training, inference, and user interaction. How to draft actor-aligned claims that reach trainers or users of the trained model.
More From Our Blog
Further articles published on our blog at blog.ai-patents.com
How to Patent Agentic AI Technology
As AI evolves from passive predictive models into agentic systems capable of planning, tool use, multi-step workflows, and autonomous decision-making, patent protection strategies must evolve too.
Beyond Generic AI: How to Patent Real Machine-Learning Innovation After Recentive
Advances in AI have triggered a surge in patent filings, but not all AI-driven ideas qualify as patentable inventions — especially after recent legal precedents.
How to Successfully Protect AI Technology in 2025: Part II
The Legal Framework — Understanding Section 101 Patent Eligibility and navigating the complexities of AI patent prosecution.
Patent Protection vs Open Source: A Strategic Guide for Startups
The software industry presents a fundamental tension between two powerful forces: the open source movement's emphasis on collaboration and patent protection's focus on exclusivity.
Introducing the "Spec Map" — A Hierarchical Part/Feature Map for Patents & Technical Disclosures
In the world of patents and technical disclosures, precision and structure are critical. A new approach to organizing technical documentation for patent applications.
From Structured Maps to Intelligent Retrieval: A New Paradigm for Technical RAG Systems
The era of "good enough" AI for technical applications is ending. While most RAG systems rely on flat embeddings, a new approach offers unprecedented precision.