While the US patenting environment has changed dramatically in the past two years, high-tech startups still need to pursue patents—but with a different focus. Here’s how and why.

Patents are being reframed today as strategic/financial assets, not just legal shields. IP teams are being pushed beyond pure legal protection into business strategy, combining patent, trademark, licensing and competitive intelligence data to identify valuable assets and make sharper decisions about innovation investment and capital allocation. Litigation data backs this up; patents that reach court tend to have stronger claim drafting, deeper specifications, larger families and more active continuation strategies, meaning portfolio quality matters more than filing volume. While many large enterprises have used this approach for decades, today, startups are adopting the same strategy to better convey value to investors and markets.

AI is measurably moving the needle on patent value. A Harvard Business School analysis of 1.8 million patents granted from 2001 to 2023 found that AI-related inventions are worth about 9.6% more than non-AI inventions on average, after controlling for all relevant factors. It's not just the original AI developers capturing that value; companies applying AI to innovate elsewhere are seeing returns too. Separately, the share of AI-related patents granted to public companies grew from 9% to 30% between 2001 and 2023, and that trajectory has continued sharply since. (Many patent applicants elect to keep their filings secret for up to 18 months after filing, so 2025-2026 data is incomplete or unavailable.)

But "just file on AI" isn't the play anymore. One industry analysis put it bluntly: broad AI patents (e.g., on neural networks generally) are now seen as too late and too broad. Instead, value has migrated to vertical AI deeply integrated into specific industrial workflows, with the earlier wave of generic "GenAI wrapper" patents already considered dated. My experience negotiating with the USPTO in 2025-2026 confirms this. Those earlier applications are subject to persistent rejections. So AI-based patent claims need to address model improvements, graphs of agentic workflows, domain-specific decision logic that drives transitions in the agent graph, consequence-accountability-governance control techniques, or other technical steps.

The licensing and monetization mindset is spreading down-market. The largest patent-holding companies now control over $2 trillion in licensable IP, generating more revenue from licensing than many operating companies generate from their products. Notably, the structuring principles big licensors use (filing claims that cover broadly practiced technology rather than narrow implementations) are increasingly being used by startups with as few as 3-5 patents, not just billion-dollar portfolios.

PTAB/litigation dynamics are shifting leverage. Patent executives at HP, SanDisk, Renesas and IBM have flagged that changes to PTAB practice under USPTO director John Squires are making inter partes review defenses less predictable and that restricted PTAB access is lowering perceived invalidation risk. And the transaction costs inherent in pursuing federal court invalidation litigation remain sky-high. As a result, patent owners have more licensing leverage, shifting how much a granted patent is "worth" as a negotiating asset.

Consequently, planning, quality and strategic drafting are the differentiators. Your IP counsel must focus early on business alignment and budget/streamline prosecution smartly, since companies increasingly view patents as strategic assets for licensing and enforcement, not just defensive shields. Consider Track One prioritization, credibly focused, novel claims and early examiner interview engagement. Use AI-supported patent assessment tools like Clerq, which Baker Botts supports, to find and claim around the closest prior art literature.

Here’s how I can help you make this happen.

1. Let’s examine your portfolio strategy, not just filing throughput. Our clients are moving away from "file everything" toward a focus on quality and monetization. Let’s review your claim breadth for licensing potential, continuation strategy, and family depth, not just getting something allowed, and adjust your prosecution strategy accordingly. This doesn’t require a deep commitment of your company to us, but yields high value for you from our experience, precision, and technical fluency.

2. Make the outcome filing or prosecuting "licensing-ready" patents from the start, not just defensive filings. Use the recommendations we provide to modify your prosecution strategy or select which new cases to file. Use Clerq-based search results to draft claims that are confidently novel.

3. Trust our technical credentials. Baker Botts has over 150 patent professionals with 200 advanced degrees in fields ranging from databases to pharmaceuticals and everything in between. Our patent lawyers, agents, and technical advisors have industry and examiner experience. Tell us about your field and let us match professionals to your tech. In AI, we are not generalists but have experience specifically relevant to governance and control tools, software sandboxing, travel, advertising, customer data analysis, procurement, energy demand assessment and others. Let us align our vertical-specific expertise with your product focus.

4. Address the PTAB and enforcement now, even if your current goals are defensive. Our PTAB trial experience is extensive. We can advise you on how PTAB and federal court unpredictability affects drafting decisions today, starting with focused, credible claims and stronger specifications engineered to survive IPR challenges.

Baker Botts stands out as a full-service IP firm, with deep experience in preparation, prosecution, examiner negotiations, PTAB trials, and federal court enforcement. You’ll achieve the best outcomes for your portfolio by turning to a professional team that does more than draft and file. To get started, let’s talk about a portfolio review today.