DIY AI Patent Filings Can Leave Startups Legally Exposed

By Daniel Okafor

AI tools can now produce something that looks like a patent application in minutes. That speed is exactly why do-it-yourself filing has become a liability for startup founders, according to Michael Drapkin and Nathan Mutter, founding partners of Landfall IP, an AI-native patent law firm based in Boulder, Colorado.

The authors lay out the problem in a new analysis for Crunchbase News. General-purpose AI can research existing patents, document inventions and generate application drafts. But the technology's capabilities are "jagged," they write: it performs some tasks remarkably well while missing legal and strategic issues that determine a patent's commercial value.

An application can look sophisticated while providing little meaningful protection.

Why claim scope is the hidden trap?

A patent's claims define its legal boundaries — what the holder can stop competitors from doing. Mistakes involving claim scope may remain hidden until an investor, acquirer or competitor scrutinizes the company's intellectual property. By then, the damage is done.

The authors frame the central question plainly: not whether startups should use AI, but how much founders can safely do themselves and when legal expertise becomes essential.

Where does AI genuinely help?

Founders can use AI to organize technical documentation, summarize existing patents and identify potentially relevant prior art. Turning engineering notes into an organized invention disclosure, for example, can reduce preparation costs and help counsel focus on strategy.

The bigger opportunity, the authors argue, lies beyond DIY. Counsel can orchestrate AI across the workflow — coordinating invention discovery, drafting, drawings and quality checks. AI can help counsel explore alternative implementations with inventors, identify disclosure gaps and check support for claims. Attorneys direct that work, evaluate the results and connect each step to business needs.

That combination can produce more comprehensive applications with substantially greater efficiency. It is fundamentally different, the authors note, from generating an application with a chatbot and asking a lawyer to review it afterward.

When does DIY filing become a liability?

Drapkin and Mutter identify three failure modes:

• Incomplete provisional applications. A provisional's early filing date is only as useful as its disclosure. Later claims receive that date only if the provisional adequately describes and enables the claimed invention. A DIY draft may omit technical details, alternatives and variations essential to future claims — and adding those details later cannot repair the original disclosure retroactively.
• Premature disclosure. Pitching investors, demonstrating products or announcing innovations before filing can jeopardize patent rights, particularly internationally. Public disclosure can also destroy trade secret protection. Sharing confidential invention details with AI tools without evaluating their data practices creates additional risks.
• Unclear ownership. Paying an employee or contractor does not substitute for securing invention assignments. The authors advise securing those agreements when people join, not waiting for diligence to expose gaps. Founders should also assign relevant work to the company and verify obligations to prior employers or universities. Ownership gaps can derail fundraising or acquisitions.

The cheapest filing option is not necessarily the most economical. A poorly prepared application can create false confidence while leaving valuable innovations exposed.

When does counsel matter?

Counsel can help decide what to patent versus maintain as a trade secret, prioritize filings within budget and assess deadlines before disclosure limits the options. An internal process competitors cannot reverse engineer may be better kept secret.

Claim scope requires particular judgment. A fraud detection claim tied to one algorithm may leave competitors free to deliver the same functionality using another. Overly broad claims may overlap with existing technology. Valuable claims target what competitors must do to compete — ideally, a single identifiable competitor can infringe them, and that infringement can be detected and proven.

Why do investors look beyond 'patent pending'?

Filing volume says little about commercial value, the authors write. Investors and acquirers ask three questions:

• Do you own it? They want signed assignments from relevant founders, employees and contractors, with no competing claim from a prior employer or university.
• Does it protect your moat? Claims should cover what the business depends on, not a feature it will abandon next year, and the application must contain enough detail to support them.
• Can you enforce it? Claims should be difficult to design around, and infringement should be something a court can realistically see proven.

What should founders ask?

AI's greatest value is not generating more applications at lower cost, the authors conclude. It is making meaningful protection more accessible without sacrificing legal judgment. Founders should ask how a firm coordinates its AI tools and legal expertise — not simply whether it uses AI.

The most expensive patent mistake, Drapkin and Mutter warn, may be discovering too late that a DIY application never adequately protected the innovation on which the business was built.

Drapkin previously founded one of the largest electronics and computer patent prosecution practices in the U.S. at an Am Law 100 firm and was named U.S. Patent Practitioner of the Year by Managing IP in 2022. Mutter has served as top outside counsel at Oura, Coinbase, Rubrik and Salesforce.