PATENTS - California’s innovation economy depends on more than research talent and venture capital. It also depends on whether inventors can protect valuable ideas without exhausting their time and funding. Recent shifts in patent fees, artificial intelligence rules, and Patent Trial and Appeal Board procedures are changing that calculation.
For California startups, universities, investors, and established technology companies, the issue isn’t simply whether patents are stronger or weaker. It’s whether the system remains predictable enough to support long-term investment.
Patent policy changes affect companies before they reach the market
Patent policy can sound remote from everyday economic development. In practice, it influences decisions made long before a company opens an office, hires engineers, or launches a product.
An early-stage business may spend years developing a medical device, semiconductor component, battery system, or software-based process. During that period, the company may have little revenue. Its most valuable asset may be the possibility of preventing competitors from copying the invention once it reaches the market.
That possibility affects several business decisions:
• Whether investors will fund additional research.
• Whether a startup can negotiate a licensing agreement.
• Whether a university can transfer a discovery to a commercial partner.
• Whether a larger company will acquire a smaller inventor.
• Whether the business keeps an invention secret or discloses it through the patent system.
A policy change doesn’t need to eliminate patent rights to affect those decisions. Higher filing costs, uncertain eligibility rules, or a greater risk of post-grant challenges can alter the expected value of a patent.
That matters in California because the state has a high concentration of businesses that invest heavily before earning meaningful revenue. Biotechnology companies, climate technology developers, aerospace suppliers, and chip designers often operate on long research timelines. When patent protection becomes more expensive or less predictable, those companies feel the effect early.
Higher patent costs can change which inventions get protected
Patent expenses have always required planning, but recent fee adjustments have made portfolio discipline more important. The USPTO’s current patent fee schedule covers filing, examination, maintenance, extensions, continued examination, and several other actions that may arise during prosecution.
The filing fee is only one part of the cost. An application may require amendments, interviews with an examiner, responses to rejections, continuation filings, foreign applications, and maintenance payments after a patent is granted.
For a large corporation, a fee increase may be manageable across an established legal budget. A five-person startup may face a different choice. It might have enough funding to protect one core invention but not three related improvements.
That pressure can produce a more selective patent strategy. Instead of filing an application for every potentially useful development, companies may concentrate resources on inventions that are central to the product, difficult to design around, and likely to remain commercially relevant.
Selectivity isn’t necessarily harmful. Weak applications can consume money without creating useful protection. The concern is that promising inventions may go unprotected because the company lacks the resources to evaluate them properly.
This divide may become particularly important for founders outside California’s largest venture networks. A well-funded Silicon Valley company can absorb higher prosecution costs more easily than a university spinout in Riverside, a manufacturing startup in the Central Valley, or an independent inventor in Los Angeles.
Public policy discussions about innovation should therefore consider access, not just total patent activity. The number of filings tells only part of the story. Policymakers should also ask which applicants can afford to remain in the system through examination, appeals, and maintenance.
AI-assisted invention requires better records of human contribution
Artificial intelligence is now part of research and product development across many California industries. Engineers use computational systems to test designs. Pharmaceutical researchers use them to identify promising compounds. Product teams use generative tools to explore technical alternatives.
The legal question isn’t simply whether an invention involved AI. The more important question is what human beings contributed to the claimed invention.
Current U.S. patent practice still requires human inventors. That makes documentation more important when an automated system plays a substantial role in generating, testing, or refining an idea.
A company should be able to reconstruct how the invention developed. Who framed the technical problem? Who selected the relevant inputs? Who recognized that an output could solve a practical problem? Who modified the proposed design? Who decided which features belonged in the final invention?
Those questions may sound like legal housekeeping, but they also affect business value. If a company can’t establish inventorship confidently, it may encounter problems during patent examination, financing, licensing, or litigation.
California businesses using generative systems should treat invention records as part of ordinary research management. Dated design notes, laboratory records, code commits, meeting summaries, and version histories can help distinguish human technical judgment from machine-generated output.
Teams should also review confidentiality practices. Entering unpublished technical information into an external system may create risks that are separate from inventorship, including contractual, trade secret, and data governance concerns.
Because the analysis depends on the facts of each project, businesses should seek appropriate US patent guidance before relying on an AI-assisted development as the foundation of a filing strategy. The goal isn’t to avoid these tools. It’s to preserve a clear record of the human work that turned a generated possibility into an actual invention.
Post-grant review is changing the value of an issued patent
Receiving a patent doesn’t always end the dispute over validity. Certain patents can be challenged through proceedings before the Patent Trial and Appeal Board (PTAB), including inter partes review.
Recent policy changes have placed renewed attention on when the PTAB should decline to institute a proceeding. Factors may include related litigation, the timing of parallel cases, prior examination, the parties’ conduct, and broader policy considerations.
For patent owners, greater use of discretionary denial may reduce the likelihood of facing overlapping validity disputes in both federal court and the PTAB. That can make an issued patent appear more durable.
For accused infringers, the same shift may remove a forum that can resolve validity questions more quickly than district court litigation. A company facing a questionable patent may have fewer practical options for challenging it before trial.
California sits on both sides of this issue. The state contains thousands of patent owners, but it also contains companies that regularly defend against infringement claims. A biotechnology startup may want its core patent insulated from repeated challenges. A software company receiving a demand letter may want an efficient way to contest overly broad claims.
This is why simple labels such as “pro-patent” and “anti-patent” don’t capture the economic effect. A policy that strengthens one company’s bargaining position may increase another company’s litigation exposure.
Investors and corporate buyers will need to examine more than the number of patents in a portfolio. They should consider claim quality, prosecution history, possible prior art, pending litigation, and the realistic availability of post-grant review.
Uncertain patent eligibility can redirect research investment
Some of California’s most commercially important fields sit near difficult questions of patent eligibility. Software, medical diagnostics, data processing, and artificial intelligence can involve claims that courts or examiners view as abstract ideas, natural laws, or natural phenomena.
The policy problem isn’t that every discovery should qualify for a patent. Basic scientific relationships and broad concepts shouldn’t become private property merely because they’re described in technical language.
The economic problem arises when businesses can’t predict where the boundary lies. Two inventions may require similar levels of research investment yet receive different treatment because of how their claims are drafted or how a decision-maker characterizes the underlying concept.
Uncertainty influences behavior. A company may narrow its research program to focus on inventions with clearer protection. It may rely on trade secrets rather than public disclosure. Investors may discount businesses whose value depends heavily on patent eligibility in unsettled areas.
This can be especially consequential in diagnostics. Developing and validating a test may require considerable scientific work, but protecting the resulting method can be difficult when the claimed invention depends on detecting a naturally occurring relationship.
Software companies face a related challenge. A claim that merely describes a business result performed on a computer is vulnerable. A claim tied to a specific technical improvement may have a stronger position, but the dividing line isn’t always easy to apply.
Clearer standards would not guarantee that more applications receive patents. They would give applicants, competitors, and investors a better basis for estimating risk before committing capital.
California needs a broader response than filing more patents
California can’t control federal patent law, but state and local institutions can reduce the burden of uncertainty.
Universities can improve invention disclosure systems and train researchers to document human contributions when automated tools are involved. Technology transfer offices can evaluate commercial potential before spending limited funds on broad filing programs.
Incubators and economic development organizations can give founders early access to intellectual property education. Many expensive mistakes occur before a patent professional sees the project, such as public disclosure, unclear ownership agreements, incomplete inventor records, or a failure to coordinate U.S. and foreign filing plans.
Investors can also improve diligence. Counting issued patents is a poor substitute for examining what those patents cover. A smaller portfolio with commercially meaningful claims may be more valuable than a large collection of narrow or easily challenged rights.
Companies should connect patent decisions to product strategy. They should ask whether an invention will still matter when the patent issues, whether competitors can detect infringement, and whether trade secret protection would be more practical.
That doesn’t mean every startup needs an elaborate global portfolio. It means patent spending should reflect the company’s actual technology, market, funding, and competitive risks.
Predictability is the real economic issue
Patent debates often focus on whether rights are too strong or too weak. California’s innovation economy needs a more practical question: Can inventors, competitors, and investors make informed decisions before spending years and millions of dollars?
Reasonable people will disagree about fees, eligibility, AI-assisted invention, and post-grant review. But frequent shifts without clear transition rules increase costs for everyone.
California will remain a major source of new technology because of its universities, workforce, capital, and industrial depth. Keeping that advantage will require more than celebrating invention. It will require a patent system that gives businesses a reasonably clear view of what can be protected, what protection will cost, and how durable that protection is likely to be.
###
