In a sector where “patented” gets used as a marketing flourish, the single most useful literacy skill is distinguishing a granted patent from a published application. They look similar — both have document numbers, abstracts, and claims — but they are different legal animals. A grant is an issued right; a published application is a disclosure of intent. The tell is the kind code, the letter-and-digit suffix at the end of a U.S. document number: A1 marks a published application, while B1 and B2 mark issued grants (B1 where no pre-grant publication preceded it, B2 where one did). “US20240123456A1” is an application; “US12345678B2” is a granted patent. Reading the suffix first prevents the most common error in crypto-IP coverage: treating a filing as if it were a granted monopoly.

The legal weight behind that letter comes from statute. A granted patent confers the right to exclude under 35 U.S.C. 154; the patent issues under 35 U.S.C. 151 once the application is in condition for allowance and the issue fee is paid. The publication of an application, by contrast, happens under 35 U.S.C. 122(b) — generally about eighteen months after filing — and publication is not allowance. The claims in a published application are what the applicant asked for; they have not been examined to a final allowance and routinely get narrowed, split, or rejected before any patent issues, if one issues at all. Section 122(b)'s text frames publication as a disclosure event, and the prior-art statute makes the consequence explicit.

“the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.”— 35 U.S.C. § 102(a)(2), source

That quotation is doing double duty. It confirms that the statute treats “a patent issued under section 151” and “an application… published… under section 122(b)” as two separate categories — the grant and the publication are different things in the law's own words. And it shows their shared after-effect: both become prior art as of their effective filing date against later inventions. So a published application, even one that never matures into a grant, still does real work — it can block someone else's later patent. What it does not do is give its applicant the right to stop others from practicing the invention. Only the grant does that.

Why the distinction reshapes a blockchain story

Crypto and blockchain IP is filing-heavy. Companies publish large volumes of applications staking out custody, consensus, settlement, and wallet concepts long before — or instead of — obtaining grants, and the same invention can ride through a chain of continuations, each generating new published applications. If a report counts every published application as a “patent,” it inflates an estate with intent that has not been allowed. The honest count separates the two: grants (B1/B2) are enforceable coverage of allowed claims; applications (A1) are disclosed positions that may narrow or lapse. A landscape that conflates them tells you what a company wanted to claim, not what it got.

There is a second-order subtlety worth flagging, because it cuts the other way. A published application is not legally inert in the interim: 35 U.S.C. 154(d) provides for provisional rights — a possibility of a reasonable royalty — running from publication, but those rights are contingent on the application actually issuing as a patent with substantially identical claims, and on the infringer having actual notice. So “just an application” understates it slightly: a publication can become prior art immediately and can support a contingent royalty later. But the headline rule holds. The right to exclude — the thing people mean when they say a method is “patented” — attaches only on grant.

The reading habit

A related trap is the continuation family, which is common in crypto IP and inflates naive counts further. A single invention can spawn a chain of related filings — each a continuation that shares the original disclosure but pursues different claims — and each can be separately published as an application and, later, separately granted. The result is several documents, sometimes with nearly identical titles and abstracts, tracing back to one specification. In the custody and distributed-ledger authorization space, for instance, the same titled invention can appear as one assignee's earlier grant and a successor entity's later grant after a portfolio transfer, alongside continuation filings in between. Counted carelessly, that one inventive thread reads as a small estate. Read correctly, it is one disclosure with a family of claims of varying scope. The kind code tells you which members are grants versus applications; the family relationship tells you they are not independent inventions. Both reads are needed before a “number of patents” figure means anything.

So the practical routine for any blockchain patent claim is short. Read the kind code: A1 is a filing, B1/B2 is a grant. Match it to the verb: “we filed” or “we have applications pending” describes A1 documents; “we were granted” or “we own” should point to B-coded records. And when a press release or landscape figure says a company holds N blockchain patents, ask whether N counts grants, applications, or both — because under 35 U.S.C. 151 and 122(b), those are not the same number, and they do not carry the same right.