By Tom Marlow, Chief AI Officer of Black Hills AI
The convergence of artificial intelligence and patent law has shifted from academic speculation to daily practice reality. As AI systems increasingly participate in invention processes and generate vast quantities of potential prior art, patent practitioners need more than theoretical frameworks—they need battle-tested strategies from attorneys who understand both the technology and the law. Finding patent AI software for lawyers is now a top priority for firms looking to stay competitive.
That’s why I am proud to announce, as the Chief AI Officer of Black Hills AI, that I have co-authored the comprehensive AI chapter in the newly released 2025 US Patent Prosecutors Desk Reference. This isn’t just another legal treatise—it’s a practical roadmap written by patent attorneys. In this case, it includes insights from one who’s been implementing legal tech for AI patent drafting for IP professionals while navigating the complex legal landscape firsthand.
The Patent Attorney Advantage in AI Implementation
At Black Hills AI, we believe that effective AI tools for lawyers must be built by practitioners who understand the nuances of prosecution, the weight of legal precedent, and the real-world pressures facing IP professionals. Our OttoIP solution reflects this philosophy, incorporating the same legal expertise and strategic thinking that informs our thought leadership in this rapidly evolving field.
Key Challenges Patent Attorneys Face Today
The Inventorship Puzzle: Human + AI Collaboration
The Question Everyone’s Asking: When AI assists in the invention process, who qualifies as an inventor?
While Thaler v. Vidal established that AI systems cannot be inventors, the practical implications run much deeper. The new chapter explores how the “Pannu factors” apply to AI-assisted inventions, requiring each named inventor to make a “significant contribution” to conception.
Strategic Takeaways:
- Constructing prompts “in view of a specific problem to elicit a particular solution” may constitute inventorship
- Developing “essential building blocks” for AI systems can establish inventorship rights
- Taking AI output and making substantial modifications can transform someone into a proper inventor
Patent Eligibility Under Alice/Mayo: AI Claims That Survive
The Reality: AI and machine learning innovations aren’t categorically abstract—but they require careful claim crafting.
Modern AI legal assistants can help streamline this process, but the legal strategy remains paramount. Recent PTAB decisions like Ex Parte Holtmann-Rice and Ex Parte Martin demonstrate how AI-related claims can overcome § 101 challenges by:
- Demonstrating concrete technological improvements (enhanced memory usage, improved accuracy metrics)
- Integrating abstract concepts into practical applications with specific implementations
- Avoiding results-oriented functional language that courts view skeptically
As the Federal Circuit recognized in XY, LLC v. Trans Ova Genetics, improvements using mathematical algorithms remain patent-eligible when “the invention lies in the improvement.”
The Prior Art Flood: When AI Generates Everything
The Emerging Challenge: AI systems can produce enormous volumes of content that may constitute prior art, but traditional analysis frameworks weren’t designed for algorithmically-generated references.
Strategic Defense Approaches:
- Public Accessibility Challenges: Argue that AI-generated references lack adequate “roadmaps” for skilled artisans to locate them. Leverage precedent from In re Cronyn and Blue Calypso v. Groupon regarding accessibility requirements.
- Technical Deficiencies: Demonstrate that AI-generated disclosures contain technical errors or insufficient implementation details.
Key Insight: “AI generated prior art differs fundamentally from traditional methods in that generated prior art is produced through language algorithms rather than technical expertise or scientific knowledge.”
Written Description and Enablement: Avoiding the Functional Claiming Trap
The Danger Zone: AI patents face heightened scrutiny under 35 U.S.C. § 112, where broad functional claiming often proves fatal.
Critical Lessons from Recent Decisions:
- Avoid the “Black Box” Problem: Claims reciting broad terms like “machine learning component” without structural support risk indefiniteness rejections.
- Provide Commensurate Enablement: The In re Starrett decision rejecting claims encompassing 140 trillion embodiments serves as a cautionary tale.
- Leverage Structural Terminology: Use recognized structural terms linked to specific circuitry types to avoid automatic means-plus-function treatment.
Looking Ahead: The Evolving PHOSITA Standard
As AI tools for lawyers become standard equipment for practitioners, courts may begin considering whether these capabilities have reshaped the “person having ordinary skill in the art” (PHOSITA) standard. While no direct case law yet addresses this shift, the KSR framework emphasizing “ordinary creativity” and “predictable results” provides adaptable benchmarks for future developments.
Patent Attorney-Backed AI Solutions
This deep understanding of both AI technology and patent law fundamentals drives everything we do at Black Hills AI. Our OttoIP platform doesn’t just automate tasks; it acts as one of the premier AI legal assistants, incorporating the strategic thinking and legal expertise that patent attorneys need to navigate these complex AI patent challenges effectively.
Whether you’re engaging in AI patent drafting, challenging algorithmic prior art, or crafting claims that survive § 101 scrutiny, our tools reflect real-world prosecution experience combined with cutting-edge AI capabilities.
Resources for Patent Practitioners
Get the Complete Analysis: The 2025 US Patent Prosecutors Desk Reference is now available and includes Thomas Marlow’s comprehensive chapter on AI implications for patent practice. This essential resource provides detailed frameworks and strategic guidance for modern IP practice.
Learn More About Our Approach: Join our upcoming webinars where we dive deeper into practical applications of AI in patent practice, demonstrating how AI software for lawyers is shaped by patent attorney expertise.
Ready to Discuss Your AI Strategy? Contact Black Hills AI to explore how legal tech for AI patent drafting can enhance your practice while maintaining the compliance and transparency standards that IP professionals demand.
Thomas Marlow is Chief AI Officer of Black Hills AI and a contributing author to the 2025 US Patent Prosecutors Desk Reference. For speaking engagements, CLE presentations, or consultation on AI-related patent matters, contact Black Hills AI.