How to Overcome a Section 101 Rejection in 2026: A Practical Guide to Subject Matter Eligibility Declarations

For a decade, a Section 101 rejection was the most frustrating letter a software, AI, or diagnostics company could get from the US Patent and Trademark Office (USPTO). The invention might be new, non-obvious, and commercially important, yet the examiner would call it an abstract idea, and the back-and-forth that followed was mostly lawyers trading legal arguments.

That dynamic is changing. Since December 2025, the USPTO has actively encouraged applicants to answer eligibility rejections with evidence, through a tool called a Subject Matter Eligibility Declaration, or SMED. The Office has refined the guidance twice since — most recently on September 29, 2026 — and says early users are having success. If your portfolio includes software, AI/ML, data analytics, or diagnostics, this guide explains what SMEDs are, how to use them well, and the risks to manage.

Why eligibility has been so hard

Under 35 U.S.C. Section 101 and the Supreme Court’s Alice framework, a claim directed to an abstract idea, law of nature, or natural phenomenon is ineligible unless it contains something more — typically, a specific technical improvement. The problem has always been proof. Whether a claimed method actually improves how a computer, a model, or a diagnostic assay works is often a factual question, yet the debate in prosecution tended to be purely legal, with examiners and applicants talking past each other.

The current USPTO leadership has signalled a more inclusive approach to eligibility for emerging technologies, as we described in The State of US IP Strategy in 2026. SMEDs are one of the most concrete ways that policy reaches individual applications.

What a SMED is

A Subject Matter Eligibility Declaration is a voluntary declaration under 37 CFR 1.132 (Rule 132) — the same rule long used to submit evidence on obviousness, such as test data or commercial success. A SMED uses that mechanism to put facts on the record about eligibility: what technical problem the invention solves, how the claimed features produce a technical improvement, and how that differs from prior approaches.

How the guidance has evolved

• December 4, 2025: the Director issued memoranda launching the current SMED programme, reminding examiners that applicants may submit eligibility evidence and setting out initial best practices.
• April 30, 2026: a superseding best-practices memo reaffirmed the approach, described the guidance as a living document, strongly recommended filing SMEDs separately from declarations on other grounds, and reported that early voluntary submitters were having success. It also noted practitioner feedback questioning whether examiners need to weigh an unrebutted SMED at all, and said the Office was considering an MPEP update on that point.
• September 29, 2026: an updated memorandum supplemented the April guidance with a flowchart for reviewing SMEDs and further best practices on timing, required statements, and how examiners evaluate the evidence.

Best practices for a SMED that works

Keep it standalone and eligibility-only

A single Rule 132 declaration can technically address several rejections, but the USPTO recommends a separate SMED focused solely on Section 101. Mixing eligibility evidence with obviousness or written-description arguments makes it harder for the examiner to isolate what matters and weakens both submissions.

Tie every fact to the claims

The declaration’s value lies in facts that connect to the claim language. Explain which claimed elements produce which technical effect. General statements about the importance of the field, or legal conclusions dressed up as facts, add little.

Stay inside your specification

A SMED cannot add new matter or cure gaps in the disclosure. It explains and evidences what the application already describes. If the specification never mentions the technical improvement you now want to rely on, the declaration cannot supply it — which is a drafting lesson for future filings as much as a prosecution tactic.

Use concrete evidence where you have it

Benchmarks, latency or memory measurements, accuracy comparisons, reduced compute cost, improved assay sensitivity — quantified technical results are the most persuasive content. Where data is unavailable, a qualified technical expert’s explanation of why the claimed arrangement produces the improvement can still carry weight.

Consider filing early

A SMED does not have to wait for a rejection. Filing proactively, in anticipation of eligibility concerns, can let you frame the analysis before the examiner settles on a position — particularly in art units where Section 101 rejections are routine.

The risk most applicants overlook

A SMED is sworn evidence that becomes part of the permanent prosecution history. If the patent is later asserted, the accused infringer can depose the person who signed it and test every factual statement under cross-examination. Overstated claims of improvement can also be used to argue claim scope or to attack credibility at trial.

When a SMED makes sense

Situation	SMED worth considering?	Why
AI/ML method with measurable performance gains	Strongly	Quantified technical improvement is exactly what the evidence is for
Diagnostic method with improved sensitivity or specificity	Strongly	Lab data can rebut a law-of-nature characterisation
Business-process software with no technical effect described	Rarely	A declaration cannot add an improvement the specification lacks
Repeated 101 rejections in a crowded art unit	Often	Forces engagement with facts rather than another round of argument
Patent likely to be asserted in litigation	Carefully	The declarant will be deposed; precision matters more than breadth

What this means for your portfolio

For founders and in-house teams, SMEDs change the economics of eligibility. Applications that would once have been abandoned after a third Section 101 rejection may now be worth pursuing, and inventions previously written off as unpatentable deserve a second look. The investment shifts from arguing to documenting: capturing benchmark data and technical rationale while the work is being done, not reconstructing it years later.

• ☐  List pending applications with outstanding or likely Section 101 rejections.
• ☐  Identify which ones describe a concrete technical improvement in the specification.
• ☐  Gather supporting data: benchmarks, test results, and comparisons with prior approaches.
• ☐  Select a credible declarant who can defend the evidence long term.
• ☐  Revisit recently abandoned software, AI, and diagnostics applications that may be revivable or worth refiling.
• ☐  Update invention-disclosure forms so engineers record measurable technical improvements at the time of invention.

Remember that eligibility is only one hurdle. With the PTAB now less available to challengers and US patents generally more valuable as leverage, a granted patent with a clean, credible prosecution history is worth more than it was a few years ago.

The bottom line

Subject Matter Eligibility Declarations give applicants something they have lacked since Alice: a structured way to put facts in front of the examiner on eligibility. Used well — standalone, claim-focused, grounded in the specification, and signed by someone who can defend it — a SMED can move a stalled application forward. Used carelessly, it creates a record that an infringer will mine later. If you have AI, software, or diagnostics applications stuck on Section 101, a targeted SMED strategy review is a high-return place to start, and it is work we do regularly for growth-stage clients.

---

Disclaimer: This article provides general information about IP strategy and should not be relied upon as legal advice. IP laws vary significantly by jurisdiction and every business situation is unique. Consult qualified counsel about your specific circumstances.