June 2, 2026 - Patent-$tartup

Most inventors think their patent is great. They are wrong — and they do not know it. The prior art that kills a patent does not live only in the United States. It lives in Europe, China, India, and every other major patent office in the world. You will not find it unless you search.

That is why I tell every client the same thing: spend five hundred to eight hundred dollars on a prior art search before you spend ten thousand dollars on a patent. With AI-assisted search software, the cost of a search has come down from over $3,000 to less than $1,000 — making it one of the highest-ROI decisions an inventor can make.

Prior Art Is Everywhere — Not Just the US

The patent office does not just look at US patents when deciding whether your idea is new. It looks at everything published anywhere in the world, in any language. That includes patents from Europe, China, India, Japan, and Korea — and every paper, product manual, and conference proceeding ever published.

When I run a search for a client, I search the same major patent offices: US, Europe, China, India. Almost every time, we find relevant prior art. Almost every time, the inventor had no idea it existed.

What a Patent Search Tells You: Strong, Average, or Weak

After a search, the result falls into one of three categories. Strong: the idea is novel, file with confidence. Average: there is some prior art, but we can draft around it. Weak: there is a patent or product that is already too close. You need to rethink the invention.

Even if the result is weak, the search is not wasted. We look at what is out there and figure out what novel feature you can build on top — something the prior art does not have. That feature becomes the new core of your claim.

The Real Math: $800 Search vs. $10,000+ Patent

Here are the numbers. A patentability search costs about five hundred to eight hundred dollars depending on the complexity of the invention. A full utility patent — drafting, filing, prosecution, allowance, and the first maintenance fee — runs ten thousand dollars and up, as you can see in the graph below.

For a solo inventor or a startup with limited capital, that is serious money. The search is the cheapest decision you will ever make to find out whether the ten thousand dollars is worth spending.

What an Actual Prior Art Search Report Contains

A search report has five sections, but the most important parts are Sections 3, 4, and 5. To illustrate, I have included a sample search report from one of my own patents (U.S. Patent No. 11,944,568 and its international counterpart WO 2022/082076, both directed to a non-invasive cooling system for crystallizing subcutaneous fat). Because the patent is mine, no client confidentiality is disclosed — what you see is the actual unredacted deliverable a client receives.

👉 View the full sample search report (Fat-Loss Cooling System)

Section 3 — Prior Art References Reviewed. Roughly the ten most relevant references found in the search, each with the patent number, the assignee, a similarity score from 1 to 100, and the title. Below each reference is a write-up with the relevant text and the differences from the invention. In the sample, the highest-scoring third-party references are AbbVie’s cooling-cup applicator (US 2021/0045912 A1) and Zeltiq’s curved-applicator cryolipolysis device (US 2014/0364841 A1), both scoring 85/100.

Section 4 — Feature-by-Feature Patentability Analysis. Every novel feature is marked as Likely Novel (not specifically disclosed in prior art), Weaker (partially covered), or Anticipated (substantially covered). In the sample, of the nine novel features analyzed, five were marked Likely Novel, three were Weaker, and one was Anticipated.

Section 5 — Patentability Conclusions and Filing Recommendation. Walks through the two or three most relevant prior art references, what the patent office is likely to do with your application, and whether to file a provisional or non-provisional. For first-time inventors, a provisional is almost always the right starting point — twelve months to develop the invention before committing to the full non-provisional cost.

Other patent attorneys and agents use different report structures, but the substance is similar: a clear, evidence-based assessment of whether your invention is novel against the global prior art.

Bottom line: the search gives you the facts about the novelty of your invention so you can decide whether it is worth filing.

Take the Next Step

I am Sung Oh, a USPTO-registered patent agent (Reg. No. 45,583). If you are an inventor or startup founder thinking about filing a patent — or you have already drafted an application with AI — that is exactly what our Hybrid Model is built for. You draft your application with AI, we revise it, and you end up with a quality patent at a lower cost.

Visit patentstartup.com to learn more, or send your invention notes to sung@patentstartup.com for a free initial review.

——————————————————————————————

References and Resources

[1] Sample Search Report (Fat-Loss Cooling System) — full PDF: https://drive.google.com/file/d/1PjV67OuT9US4KvzxRs80XDdWsi_BGzJw/view?usp=drive_link

——————————————————————————————

Disclaimer: I am a USPTO-registered patent agent (Reg. No. 45,583), not your attorney. This article is for educational purposes only and is not legal advice. Patent law is fact-specific; consult a qualified practitioner about your specific invention.

Seventy-seven percent of AI patent applications get rejected with a Section 101 “abstract idea” objection. By comparison, the software tech group sits at roughly 21 percent, and mechanical inventions at under 10 percent. [1]

The deck is stacked against AI inventors. But there is a way to win — and it comes down to how you draft your patent specification and your claims.

Where the 77% Rejection Rate Comes From

The 77 percent figure is from Voice of IP’s 2024 analysis of USPTO Office Action data, looking specifically at Working Group 2120 — the tech group that examines artificial intelligence, machine learning, and simulation/modeling applications. [1] That figure is more than double the 2022 rate, and roughly twenty points above the pre-2019 baseline.

Why is AI hit so hard? Because Section 101 of the Patent Act says abstract ideas cannot be patented. [2] When an examiner looks at an AI application, the surface impression is “math on a computer” — and math is the textbook example of an abstract idea. The examiner reaches for a Section 101 rejection because it is the path of least resistance.

How We Got Here: A Short History of Section 101

The pivot was 2014. In Alice Corp. v. CLS Bank International, the Supreme Court handed the USPTO a two-step test for figuring out whether an invention crosses the abstract-idea line. [3] After Alice, rejection rates for software and AI exploded. See the graph below.

In 2019, the USPTO issued the Revised Patent Eligibility Guidance (PEG). [4] That helped — briefly. Then AI took off, applications flooded in, and Section 101 rejections climbed back up to where they are today.

On August 4, 2025, the USPTO issued what is now called the Squires memo, raising the bar for Section 101 rejections. [5] Examiners can no longer reflexively call something abstract at step one. If you draft the right way, your odds change dramatically.

The Alice Two-Step, Explained

Alice has two questions. Step one: is the claim directed to an abstract idea? If the answer is no, the claim is eligible — you are done. If the answer is yes, you move to step two. Step two: does the claim add significantly more than the abstract idea itself? If yes, the claim is eligible. If no, the claim gets rejected under Section 101. See the diagram below.

Two questions. Three possible outcomes. Two of the three paths lead to an eligible claim. Only one path — “yes” at step one and “no” at step two — gets you rejected.

And here is what most inventors get wrong: being called “abstract” at step one does not kill your patent. It just sends you to step two. Most well-drafted AI patents are abstract at step one. They survive at step two by giving the examiner something concrete to point to. That is why your specification matters.

The Three Pillars of a §101-Proof Specification

Pillar 1: Anchor to a Technical Problem

Do not describe what your AI does — describe the technical problem it solves. “Predicts loan risk” sounds abstract. “Reduces GPU memory by 40 percent while preserving gradient stability during backpropagation” sounds technical. The first will get a Section 101 rejection. The second probably will not.

Pillar 2: Architectural Specificity

Do not draft a black box. Disclose the architecture: the layer types, the attention mechanism, the loss function, the training recipe. “A neural network trained on data” loses. “A transformer with sparse attention reducing complexity from N-squared to N-log-N” wins. Specificity is eligibility.

Pillar 3: At Least One Non-Abstract Step

Your independent claim needs at least one step that a human cannot do in their head. A sensor input. An actuator output. A specific data-transmission step. A forward pass written in architectural terms. Under the 2025 Squires memo, the examiner cannot say “a human could do this in their mind” anymore — as long as you give them something concrete to point to.

The Three Levels of an AI Patent Claim: Weak, Better, Best

Once your specification clears Section 101, the next question is how to draft the claim. Three levels, illustrated with a hypothetical AI fraud detector that is 99 percent accurate in testing.

Weak: “a detection accuracy of 99 percent.” A competitor ships 98.5 percent and walks around the claim.

Better: “an accuracy between 90 percent and 99 percent.” A competitor ships 99.1 percent and walks around the claim.

Best: “an accuracy of at least 90 percent.” Anything 90 percent or higher infringes. Same invention, three very different fences. The best claim removes the ceiling — but only as far down as the prior art allows.

Why AI Drafting Tools Make All of This Worse

If you let an AI tool draft your patent application, three things go wrong. The specification reads abstract because the AI describes outputs, not mechanisms — that is a Section 101 rejection waiting to happen. The claim comes out weak because the AI writes specs like a product manager. And the AI cannot balance scope against the prior art, because it does not know your prior art.

Three judgment calls in a row — and AI does not make any of them well. That is why an experienced patent attorney or agent earns their fee on AI applications, especially right now.

Take the Next Step

Visit patentstartup.com to learn more, or send your invention notes to sung@patentstartup.com for a free initial review.

——————————————————————————————

References and Further Reading

[1] Voice of IP, “BREAKING: Section 101 Rejections Soar at USPTO; 77% of AI Tech Group’s OAs Include 101 Rejections” (2024). Analysis of USPTO Office Action data, Working Group 2120 (AI + Simulation/Modeling). Available at: https://www.voiceofip.com/p/breaking-section-101-rejections-soar

[2] 35 U.S.C. § 101 (Inventions patentable). The statutory basis for subject-matter eligibility. Available at: https://www.uspto.gov/web/offices/pac/mpep/s2106.html

[3] Alice Corp. Pty. Ltd. v. CLS Bank International, 573 U.S. 208 (2014). The Supreme Court case establishing the two-step test for patent-eligible subject matter. Available at: https://supreme.justia.com/cases/federal/us/573/208/

[4] USPTO 2019 Revised Patent Subject Matter Eligibility Guidance (“2019 PEG”), 84 Fed. Reg. 50 (Jan. 7, 2019). Refined the Alice/Mayo test and added Step 2A Prong Two. Available at: https://www.federalregister.gov/documents/2019/01/07/2018-28282/2019-revised-patent-subject-matter-eligibility-guidance

[5] USPTO, “Subject Matter Eligibility Examination” memorandum (the “Squires memo”), August 4, 2025. Raised the bar for Section 101 rejections, particularly in AI/software. Available at: https://www.uspto.gov/sites/default/files/documents/memo-101-20250804.pdf

——————————————————————————————

Most inventors have no idea whether the patent they are about to file is strong or weak — until it is too late. They pay ten thousand dollars or more to draft, file, and prosecute an application, only to discover that competitors can design around the claims with ease, or that the examiner rejects the claims because they collide with prior art the inventor never knew existed.

In this article, I will walk you through the single rule that separates a strong patent claim from a weak one. I will use the same example I use with my own clients — Ford and Harley-Davidson — and explain why so many AI-drafted patent applications fall apart in prosecution.

What a Patent Claim Actually Is

A patent has two parts: the specification and the claims. The specification tells the story of your invention — what problem it solves, how it works, the embodiments you considered. The claims are different. The claims are what you actually own.

A claim is a single sentence. Inside that sentence, you list every element your invention requires. If a competitor’s product contains every single one of those elements, the competitor infringes your patent. If their product is missing even one element, they walk away.

That is the whole game. Now let us see why it matters.

The Ford vs. Harley-Davidson Example

Imagine Henry Ford files a patent on the Model T in 1908. His claim covers three elements: an engine, a transmission, and four wheels. A few years later, Harley-Davidson files a patent on the motorcycle: engine, transmission, and two wheels.

Does the motorcycle infringe Ford’s patent? The answer is NO.

Ford’s claim requires four wheels. The motorcycle has only two. Because every element of Ford’s claim is not present in the motorcycle, there is no infringement. Both companies are free to sell their products.

The Reversal: Flipping the Fact Pattern

Now flip the order. Suppose Harley-Davidson files the motorcycle patent first, with just two wheels. Later, Ford files his patent on the Model T with four wheels.

Does the Model T infringe Harley’s patent? The answer is YES — but how, if the Model T has four wheels and not two?

Here is the trick. The Model T contains two wheels plus two more. It still has two wheels — and we do not care about the extra two, because Harley’s claim only requires two. Every element of Harley’s claim is present in the Model T, so the Model T infringes. A six-wheel truck would infringe for the same reason — it still contains two wheels. Extra wheels do not save you.

This is the rule that surprises every first-time inventor: fewer elements in your claim equals broader protection. Each element you add is another loophole a competitor can use to walk around your patent.

The Rule: Less Is More — But Only To a Point

So the rule is simple: less is more. A strong patent claim contains the fewest possible elements that still describe the inventive concept. The fewer elements you require, the more products fall inside your fence.

But you cannot push this rule to the extreme. If your claim is too broad — for example, “a vehicle with wheels” — the examiner will reject it, because the prior art already covers a vehicle with wheels (every bicycle, car, and wagon ever made). Your claim has to be narrow enough to be new, but no narrower than that.

The art of patent claim drafting is finding the narrow gap: broad enough to be valuable, but not so broad it collides with the prior art. This is judgment work. It takes an experienced patent attorney or agent who has read the prior art in your field and knows exactly how far to push before the examiner pushes back.

Why AI Drafting Tools Write Weak Patent Claims

I have reviewed dozens of patent applications drafted by inventors using AI tools. Almost every single one has the same problem: the claims are too narrow.

Here is why. AI tools read your product description and dump every feature into the claim. Ten, fifteen, twenty elements. To the inventor, this looks thorough. To a patent attorney or agent, this looks like a claim that protects nothing — because a competitor only needs to skip one element to escape the claim.

AI tools also do not know the prior art. They cannot tell you which elements are truly essential to novelty, and which elements you can strip out without losing patentability. That judgment requires reading the prior art in your specific technical field. AI does not do that — at least not yet.

The result is predictable: the patent gets allowed (because narrow claims clear the prior art easily), but the patent does not protect anything. A competitor changes one feature and your patent is worthless. You spent ten thousand dollars for a piece of paper.

The Bottom Line

A strong patent claim is a balanced claim. Wide enough to capture serious competitors. Narrow enough to clear the prior art. The fewest elements possible, each one truly essential to the invention.

Getting that balance right is the difference between a patent that builds a business and a patent that builds a paperweight. It is also the difference an experienced patent professional makes — someone who has read the prior art, drafted hundreds of claims, and litigated patents in the real world.

Take the Next Step

Visit patentstartup.com to learn more, or send your invention notes to sung@patentstartup.com for a free initial review.

——————————————————————————————