What Are the Best Practices for a PTAB Appeal in 2026?
The most effective PTAB appeal is not a rehearing of every exchange that occurred below. It is a disciplined appeal that tells the Federal Circuit exactly where the PTAB made a legal or evidentiary error, ties that error to a changed outcome, and provides a clean alternative without asking the court to retry the entire case. As of September 25, 2026, that remains true even as patent eligibility policy for AI inventions and PTAB precedent continue to attract attention. A successful appellant normally focuses on claim construction, unsupported fact findings, improper evidentiary treatment, or a failure to apply the governing decision as written. The appellant should also preserve an important judgment about a challenged claim, because reversal of an entire claim is not necessary if the remaining claims support a valuable settlement. The patentee should use a different structure: preserve the broadest independent claim, concede narrower positions that were already lost, and emphasize why a sufficiently broad claim remains above a statutory threshold. In either direction, the best practice is early issue selection, a verified record, and a realistic assessment of the litigation’s commercial value.
Also worth reading: What are the core AI patent appeal success factors when challenging USPTO rejections at the PTAB and Federal Circuit? · What are the definitive best practices for AI patent prosecution in 2026 to ensure claim validity and avoid disclosure risks? · How is AI patent examination handled in 2026, and what should inventors know about the current USPTO guidelines and global practices?
A good appeal also recognizes what the Federal Circuit cannot do. It reviews the PTAB’s written decision and existing record, but it generally does not hear new evidence or revisit arguments first raised in a petition for internal review. The appeal therefore depends on the record assembled during the proceeding, including the institution decision and trial record. Treating an appeal as a new trial produces lengthy briefs, unreliable statements, and arguments that are late. The central question is whether an identified error, read against the rest of the record, would produce a different result for at least one claim. That formulation keeps the appeal connected to the remedy the client actually needs and makes it easier for counsel to edit away material that does not change the outcome.
Which PTAB Orders Can Be Appealed, and When?
Under 35 U.S.C. § 141(a), a party dissatisfied with a final written decision of the Board may appeal to the United States Court of Appeals for the Federal Circuit. A notice of appeal ordinarily must be filed within 63 days after entry of the decision. Until the decision becomes final, dissatisfied parties generally proceed through petition for internal review, potentially followed by a request for panel review, rather than an immediate merits appeal. A request for internal review must normally be filed within 14 days after the final written decision, and any request for panel review within 14 days after the Board’s response to the internal-review petition. These sequential deadlines mean that a party should not conserve an appeal argument only for the final decision if the first challenge concerns a claim the PTAB refused to institute.
Trial conduct creates another set of firm limits. A patent owner generally may not amend a claim in response to a decision on institution, and 35 U.S.C. § 326(e) bars certain claim amendments after a written decision. A party that waits because an early, unresolved issue may disappear by final decision should not assume it has preserved the issue; the clean route is to raise it promptly through the procedure then available. Counsel should calendar all dates from the PTAB’s system as well as independently verify them, because notices, weekend calculations, and later proceedings can create accidental deadline problems. A notice of appeal does not stay the PTAB’s 12-month deadline for a final written decision, so the appellant should be prepared to move quickly at both the trial and appellate stages.
How Should an Appellant Build the Record for Federal Circuit Review?
Begin with an error inventory rather than a summary of the proceeding. For each challenged construction, ask whether the PTAB adopted a definition missing from the specification, mixed claim limitations into the construction, or used a construction whose meaning would not change the result. For factual findings, identify unsupported inferences, departures from the burden of proof, improperly excluded evidence, or reliance on credibility determinations that the record does not support. For evidentiary issues, separate a genuine error from a dispute over how much weight to assign. The Federal Circuit will not substitute its own view merely because another factfinder could have resolved the dispute differently. This makes precision more useful than volume: two clearly stated reversible errors are generally better than a catalogue of minor complaints.
A strong appellate brief maps the record to the requested remedy. Use short claim charts that distinguish the proposed construction, the PTAB’s construction, the key evidence, and the effect of the error on the challenged claim. Record citations should allow the court to test the argument without reconstructing the case, and quotations should be selective rather than decorative. The appellant should address adverse decisions directly rather than assume the court will discover them, while the patentee should explain why those decisions do not undermine the narrow record-supported construction that the court should affirm. Appellate counsel should also examine whether a settlement, claim-redesignation position, or remaining independent claim makes a marginal reversal commercially less valuable than a clean, enforceable claim.
How Should a Patentee Respond Without Creating New Weaknesses?
The patentee’s appeal ordinarily starts by deciding which claims still matter. If a losing claim is a properly narrowed fallback that can survive if the broader independent claim is reversed, the appeal should explain the fallback and show that each limitation remains supported by the specification and evidence. The patentee should resist the temptation to reopen every dispute that appeared reasonable during the trial. A narrow concession can make the appellate analysis easier, especially where the Federal Circuit’s case is framed around whether the PTAB correctly construed the broadest claim. The appellant must still give the court a complete reason to reverse the claims it targets, however, because silence can be interpreted as acceptance of the PTAB position.
A patentee must also separate the merits of the challenged claim from the commercial importance of the technology. A claim may be broader than a commercially necessary feature, while a dependent claim may cover a product actually sold or expected to be sold. Before spending heavily on an appeal, counsel should compare the projected value of the claim with the expected Federal Circuit cost and the time needed to litigate the remaining patents. If the client will continue negotiating across a larger portfolio, a reversible claim construction may be more valuable than a complete victory on a nonessential theory. This evaluation should be refreshed after the Federal Circuit’s case management conference and after any settlement offer, because appellate strategy and settlement leverage can change.
IPR or PGR: Which PTAB Route Fits the Objective?
Inter partes review and post-grant review address different situations and should not be chosen simply because one has a lower fee. IPR is a relatively fast trial of the constitutionality of a patent, commonly directed at claims challenged under 35 U.S.C. §§ 102 and 103 after the America Invents Act. PGR can address broader validity and eligibility issues, but it is available only during the nine-month period following issuance. The standard fee remains $20,000 for an instituted IPR and $40,000 for an instituted PGR, subject to the USPTO’s current fee schedule and the applicable excess-claim fees. Legal fees, technical experts, and litigation support are additional and should be quoted separately rather than hidden inside a single estimate.
| Feature | Inter partes review | Post-grant Review |
|---|---|---|
| Primary timing window | After publication; a § 311(b) petition generally must be filed within 1 year of service of the infringement complaint, with limited exceptions | Within 9 months after patent issuance |
| Common grounds | Patentability under 35 U.S.C. §§ 102 and 103 | Patentability and patent-eligibility issues, including 35 U.S.C. § 101 |
| Base USPTO fee | $20,000 per instituted proceeding | $40,000 per instituted proceeding |
| Typical strategic use | Fast, claim-focused challenge with a practical timing deadline | Broader, early challenge after a new patent issues |
| Main planning caution | The one-year bar can be absolute when the qualifying petition is served | The nine-month window closes even if prior validity information is still developing |
How Do AI-Patent Appeals Change the Brief?
AI-related patents still present ordinary claim-construction and obviousness questions, but eligibility issues can add an expensive layer. The Federal Circuit applies the Supreme Court’s Alice framework: a court asks whether the claims are directed to a judicial exception and, if so, whether the elements individually and as an ordered combination add an inventive concept. The PTAB can consider § 101 along with other issues, but the appellant should not assume that every technical reference to a model, software routine, or computer improvement is enough to decide eligibility. A 2026 appeal should identify the specific claim language, the relevant specification disclosures, and the factual evidence before the Board rather than rely on a general claim that AI is important. Technical experts should be used where the dispute concerns the architecture, training, inference, memory, or technical operation, but the argument should be translated into the legal test.
The patentee’s response should make the nonconventional technical work visible. Counsel can connect a disputed limitation to a problem in the prior art, an improvement in system behavior, or a feature that the specification describes as solving a technical problem, while avoiding the argument that any novel machine implementation is automatically eligible. The appellant should identify where the PTAB converted a factual question into a categorical legal conclusion, and should explain why any factual finding about the claimed invention or the prior art was wrong. This is especially important in appeals where the AI technology is described using terms that lack an established legal meaning. A declaration from a qualified engineer can help, but the brief should show exactly which finding the declaration supports.
What Are the Most Common PTAB Appeal Mistakes?
The first mistake is asking the Federal Circuit to substitute a preferred outcome for an error. The court reviews the PTAB written decision, not the lawyer’s ideal record, so a disagreement over weight is not enough. The second mistake is treating a factfinder’s credibility determination as reversible merely because the PTAB could have resolved the issue differently. A useful appeal asks whether the record supports the finding as a whole, whether a key inference was contrary to the evidence, or whether a legal rule was applied incorrectly. The third mistake is overlooking the original petition deadline or the 63-day notice period, which can eliminate the most persuasive argument before it is heard.
Another common error is including secondary arguments that dilute the primary one. Claim construction, statutory eligibility, evidentiary admissibility, and procedural sufficiency can interact, but they should be organized so the court can see which error changes the result. Parties also fail by predicting the Federal Circuit’s reaction without accounting for the board member most likely to review the issue. A focused internal critique should identify the panel, the procedural history, and the likely appellate posture rather than assume that every PTAB decision receives uniform treatment. The final error is spending as if total victory is required when the commercial goal is a narrowed, valid claim. Settlement posture should be reassessed at filing, after briefing, and at key management conferences, not treated as an afterthought.
When Should a Party Act, and What Does It Cost?
A qualifying IPR petition generally must be served within one year after service of a complaint for patent infringement. That deadline is a starting point for case strategy, not a reason to file an underdeveloped petition. The party should identify the challenged claims, complete a claim chart, obtain technical support, and decide whether settlement remains preferable before spending on the trial. In a PGR, the nine-month issuance window matters, but preparation should begin earlier because the patentee is still gathering information. In either case, a party that has already lost a motion need not restart blindly: it should determine whether the available proceeding targets the same patent, the same claim, and the same legal defect.
Cost planning should distinguish USPTO fees from market fees. Budget $20,000 for a standard IPR and $40,000 for a standard PGR, while confirming current USPTO rates before payment and adding excess-claim charges where applicable. Attorney and expert work is a separate procurement decision, with the total depending on claim count, the number of trippers, the need for technical experts, and the expected duration of Federal Circuit review. A high-value portfolio dispute can justify a large budget, while a single low-value claim usually does not. The correct comparison is expected litigation value against the cost of both winning and losing at each stage, including business disruption and the possibility that a later claim can be saved.
A Practical 60- and 90-Day Plan
Days 1 through 30 should produce a case-specific decision. Download the complete PTAB record, create a claim chart, identify the final written decision issues, and separate claims that are commercially essential from claims that are merely valuable. Decide whether the available route is IPR, PGR, post-trial court review, or settlement, and verify the statutory and Patent Rules deadlines against the official notices. Counsel should then draft the strongest claim construction and the strongest alternative construction, noting the evidence supporting each. Early internal review is valuable because it exposes the argument that a cross-examining party or a federal judge will test first.
Days 31 through 60 should turn that analysis into briefs, a record, and a budget. Obtain an independent technical assessment where the dispute is technical, and check whether any expert or declaration creates a material factual or credibility problem. The team should prepare a decision memo explaining which claims to target, which concessions improve the appeal, and what settlement range remains rational. For an appellant, that memo should include the precise appellate remedy and the claim chart; for a patentee, it should include a surviving fallback claim. Days 61 through 90 should cover final drafting, source checking, filing, and a monitoring plan for the PTAB and Federal Circuit. The best PTAB appeal practice is therefore not a single trick or a guaranteed winning brief; it is a process that aligns the legal theory, record, claim value, timing, and cost before the client commits to the next stage.