US 5991399

I'll search for docket and USPTO records for this specific patent number.

Let me search more specifically for current litigation and any 2026 appellate activity.

Let me check specifically for any 2025–2026 Federal Circuit activity or recent transfers.

US Patent 5,991,399 — Summary

Bottom line on the "CAFC 2026 docket" request: I could not find any 2026 Federal Circuit docket entry, appeal, or other 2026 appellate activity for this patent in the searches I ran. What I did find is a 2016 Federal Circuit appeal (case 16-2700) and a large body of 2015–2020 district court and PTAB activity. I'm flagging this explicitly rather than inferring a 2026 docket exists — I have no authoritative source confirming current 2026 status, and I have no live PACER/CAFC docket access to verify.

Bibliographic data (Google Patents / USPTO-derived)

Uncertainty note: The "current assignee" and "expired" statuses above are those listed on the Google Patents page; I could not independently confirm 2026 ownership or any post-2019 reassignment, and the patent text itself does not address ownership.

Abstract (verbatim)

"Secure distribution of a private key to a user's application program (also called a 'trusted player' such as a DVD player or CD-ROM player) with conditional access based on verification of the trusted player's integrity and authenticity is provided. Once validated, the trusted player uses the private key to decrypt encrypted digital content. The private key is dynamically generated, associated with specific digital content, and communicated in real-time from a server to the trusted player in a secure manner, thereby controlling access to encrypted digital content. The key is wrapped into an executable tamper resistant key module in which the key can only be used by the right trusted player as determined by the server based on user requests and payment. The key module plugs in to the trusted player and executes to validate the player and decrypt the content. The integrity of the trusted player is correlated to its ability to perform a cryptographic operation using an asymmetric key pair in a manner that is tamper resistant, thereby preventing an unencrypted copy of digital content to be made."

Plain-language overview of the independent claims

The patent has 37 claims, with independent claims 1, 19, 20, 27, 28, 29, 31, 34, 35, 36, and 37. They cluster into three families: server-side method claims, system/apparatus/CRM claims, and "build-it / receive-it" claims.

Core method family

• Claim 1 — A method for securely distributing data: (a) generate an asymmetric key pair (public + private); (b) encrypt some "predetermined data" with the generated public key; (c) build an executable tamper resistant key module identified for a selected program, where the module contains both the generated private key and the encrypted data. This is the broadest claim — it stops at building the module (no sending, no validation/decryption steps recited).

• Claim 19 — Same generating/encrypting/building steps as claim 1, but with the full end-to-end loop added: send the module to the remote system; execute it there to check (i) the integrity and authenticity of the program and (ii) the integrity of the module itself; and decrypt the encrypted data with the private key only when both the program and the module validate.

• Claim 20 — Recites the same pipeline from a server-to-client "trusted player" framing for authorizing access to selected encrypted digital content. Adds a first step of receiving a request from the trusted player for access to the selected content, then generates keys, encrypts, builds the module for that trusted player, sends it to the client, executes it to validate the trusted player and the module, and decrypts when validation succeeds.

Computer-readable-medium / apparatus / system family

• Claim 27 — A machine readable medium storing instructions that implement the claim-1 pipeline and additionally send the module to the remote system "to verify the authenticity and integrity of the program … and decrypt … when the program is validated."

• Claim 28 — A machine readable medium whose instructions execute across a plurality of processing units, covering both the server side (generate/encrypt/build/send) and the remote side (execute the module to check program integrity/authenticity and module integrity; decrypt when validated).

• Claim 29 — An apparatus: a processor plus a storage medium whose instructions generate the key pair, encrypt the predetermined data with the public key, and build the executable tamper resistant key module containing the private key and encrypted data.

• Claim 31 — A system with two parts: a first system that generates the keys, encrypts, and builds the module; and a second system that operates as a trusted player of digital content, receives the module, executes it to check the trusted player's integrity/authenticity and the module's integrity, and decrypts when both validate.

Minimal "build" and "receive" family

• Claim 34 — A method of distributing data that was encrypted by the public key of an asymmetric pair, stated in minimal form: build an executable tamper resistant key module identified for a selected program resident on a remote system (module = private key of the pair + the encrypted data) and send it to the remote system. Notably it does not recite the original key-generation or encryption steps.

• Claim 35 — The mirror-image receive-side method: receive the module; execute it to check the selected program's integrity/authenticity and the module's integrity; decrypt the data with the private key when the validations pass.

• Claim 36 — An article (machine readable medium) with instructions for the claim-34 build-and-send behavior.

• Claim 37 — An article with instructions for the claim-35 receive/execute/decrypt behavior.

Litigation / post-grant history found (for context, not 2026)

• 2016 Federal Circuit appeal 16-2700 (listed in Google Patents' litigation data) — this is the only Federal Circuit entry I found; it is from 2016, not 2026.
• IPR2017-01721 and IPR2020-00213 (Unified Patents petitioner) — both terminated/denominated as settlement. A Unified Patents news item states that on March 20, 2020 a joint motion to terminate pursuant to settlement was filed in IPR2020-00213 between Unified Patents and Honeyman Cipher Solutions LLC regarding this patent.
• Asserted 2015–2016 by Plano Encryption Technologies against banks and others; 2019–2020 by Honeyman Cipher Solutions against, e.g., Snap (1:19-cv-01547), LogMeIn (1:19-cv-01545), Groupon, Adidas (1:20-cv-00401), ASICS (1:20-cv-00402), Evernote (1:20-cv-00403), PayPal (1:20-cv-00404), Slack (1:20-cv-01076), Costco (6:20-cv-00850), Asana (3:20-cv-00928).
• Representative asserted claims identified in the Snap complaint: "at least exemplary claims 1–2, 9–11, 34" of the '399 patent.

Caveats

1. No 2026 CAFC docket confirmed. Multiple targeted searches returned no 2026 Federal Circuit activity. I cannot say one does not exist — only that I could not find authoritative evidence of it.
2. Assignments/status. The Honeyman Cipher Solutions LLC assignment (2019) and "Expired – Lifetime" status are as listed by Google Patents; I did not verify a live USPTO PatentCenter/Assignment record as of today.
3. Claim interpretations above are my plain-language paraphrases of the claim text supplied in the authoritative patent text; they are not legal claim-construction opinions.
4. One search hit rendered the number with comma grouping ("5,991,399") and one OCR'd inventor surname as "MALISZEKKI"; per your strict rule I have not auto-corrected these in quoting, but the authoritative patent text gives the inventor as Richard L. Maliszewski.

Sources: Google Patents US5991399A (patent text, claims, assignments, family); Unified Patents insights/news (IPR2017-01721, IPR2020-00213); RPX/Stanford NPE Litigation Database and the Honeyman Cipher v. Snap Delaware complaint (asserted claims); IPWatchdog patent filings roundup (2020 case list).

I'll research the litigation history for US Patent 5991399. Let me search multiple sources.

Let me research these cases in more detail.

Let me get more details on specific cases and outcomes.

Based on my research through the Google Patents litigation record for US 5,991,399, RPX Insight, the Stanford NPE Litigation Database, CourtListener, and Federal Circuit records, here is what I found.

Litigation Involving US Patent 5,991,399

Important threshold note on the patent: US 5,991,399 ("Method for securely distributing a conditional use private key to a trusted entity on a remote system") issued 1999-11-23 to Intel Corporation. It was reassigned over time to Convera Corp. (2001), H2 Partners LLC (2011), Plano Encryption Technologies, LLC (2015-04-03), and Honeyman Cipher Solutions LLC (2019-03-08). All known litigation was brought by the two NPE owners (Plano Encryption and Honeyman Cipher) — Intel itself never asserted it. Anticipated expiration was 2017-12-18.

The patent was almost always asserted together with U.S. 5,974,550 (and sometimes U.S. 6,587,858).

Campaign 1 — Plano Encryption Technologies, LLC ("PET"), 2015–2017

Affirmative infringement suits (PET as plaintiff)

Declaratory judgment actions (PET as defendant)

Federal Circuit appeal

Campaign 2 — Honeyman Cipher Solutions LLC, 2019–2020

Affirmative infringement suits (Honeyman as plaintiff)

IPR proceedings

Caveats and confidence level

• High confidence for: the two NPE owners (Plano Encryption, Honeyman Cipher); the fact that the patent was asserted in two campaign waves; the Jack Henry CAFC appeal (2016-2700, 910 F.3d 1199, reversed/remanded Dec. 7, 2018); the Groupon venue dismissal; the Asana voluntary dismissal; and both IPR settlements.
• Medium confidence for: exact filing dates of several 2015–2016 PET cases and the precise current disposition of every 2019–2020 Honeyman case (those cases were still pending in the 2020-era dockets I could retrieve, and I could not confirm later terminations).
• Unconfirmed: the defendant identities for W.D. Tex. cases 6:20-cv-00849, 6:20-cv-00851, and 6:20-cv-00854 — I obtained the case numbers and jurisdiction from the Google Patents litigation listing for US 5,991,399 but could not retrieve the party names within my search budget. Treat these as "case exists, defendant unverified."
• I found no litigation involving Intel Corporation as an asserting party, and no case involving any patent number other than 5,991,399 in this campaign (the '550 and '858 patents were co-asserted but are different patents).
• One citation discrepancy worth flagging: the Broadway National Bank opinion cites the Delaware Fidelity DJ case as "No. 1:14-cv-777," while the Google Patents record lists "1:15-cv-00777 (filed Sept. 3, 2015)." I did not auto-correct either; the 2015 filing date appears to be the correct one.

If you need PACER-verified docket sheets and final dispositions for the 2020 Honeyman cases (and the three unconfirmed W.D. Tex. defendants), those would require pulling the individual PACER dockets — I can flag the exact docket numbers to retrieve.

I'll start by checking the structured PTAB block against what the patent page itself discloses, then search for the two PTAB case numbers I see embedded in the source.

I have the first proceeding. Now let me pull the details on the 2017 filing, which the patent page lists without a petitioner name.

Both filings settled pre-institution. Let me confirm there aren't additional proceedings (e.g., a CBM, or other defendants' IPRs) I'm missing.

PTAB proceedings on file — US 5,991,399

Flag before anything else: the structured "PTAB proceedings on file" block returns no AIA trial proceedings, but that block is incomplete here. The patent page itself carries two PTAB case records, and I confirmed both against primary Board documents: IPR2017-01721 and IPR2020-00213. The likely explanation for the ODP gap is structural — neither petition was ever instituted, so neither ever became an AIA trial. Both died in the preliminary-proceeding stage on settlement. Treat the ODP "no proceedings" return as a data-scope artifact, not as evidence that no petitions were filed.

Proceedings overview

Two (2) IPR petitions have been filed against US 5,991,399. Both terminated pre-institution by settlement (2 settled / 0 active / 0 institution decisions / 0 final written decisions / 0 claims invalidated / 0 claims sustained on the merits). The bottom-line defensive posture is unusual and worth internalizing: no claim of the '399 patent has ever been adjudicated by the PTAB — not one. No claim is canceled, so a demand letter citing claims 1, 2, 9–11, or 34 is not facially dead. But the patent has also never been "hardened" by surviving an instituted trial, and no § 315(e)(2) estoppel runs against any petitioner, which means the full prior-art record (including Aucsmith and Chang/EP 0 686 906 A2) remains unadjudicated and available to you. The real defensive leverage on this patent lies elsewhere: an uncontested claim-construction fight over the word "including," and the fact that the patent expired on 2017-12-18.

IPR2020-00213 — Unified Patents, LLC v. Honeyman Cipher Solutions, LLC

• Type: Inter Partes Review
• Filed: 2019-12-13 (petition filing date as accorded by the Board; the case number carries the FY2020 series)
• Status: Verbatim from the record: "Settlement Prior to Institution of Trial 37 C.F.R. § 42.74" — i.e., terminated by joint motion before the Board ever reached an institution decision. Google Patents records it as "PTAB case IPR2020-00213 filed (Settlement)."
• Judge panel: Timothy J. Goodson (writing), Amanda F. Wieker, Ryan H. Flax
• Petition grounds: Challenged claims 1, 2, 9–11, and 34. The public record of the termination decision confirms the claim set but does not recite the specific art; the petition was never publicly adjudicated, so I will not speculate about the references. Petitioner was Unified Patents, LLC — a defensive aggregator acting for the subscriber base, not a named district-court defendant.
• Institution decision: None. The Board expressly noted: "We have not entered a decision on whether to institute an inter partes review, and the deadline for such a decision is nearly three months away."
• Final Written Decision: None. The Board stated the termination order "does not constitute a final written decision pursuant to 35 U.S.C. § 318(a)."
• Settlement / termination: Joint Motion to Terminate filed 2020-03-20; Board granted termination 2020-03-25. The parties filed a confidential "Settlement and License Agreement" (Ex. 1015). Per the Board's order, the agreement was treated as business confidential under 37 C.F.R. § 42.74(c), kept separate from the patent file, and available only to Federal Government agencies on written request or to any person on a showing of good cause under 35 U.S.C. § 317(b). The Board cited §§ 2.5 and 6.9 of the agreement in confirming a dispute resolution, so the deal included a license — but the commercial terms are confidential. Unified's own report of the settlement notes the '399 patent had been asserted against Snap, Groupon, and LogMeIn, and previously by Plano Encryption Technologies against banks and companies including State Farm, Best Buy, Etsy, Match.com, and Shutterfly (Unified Patents, 2020-03-23).
• Appeal: None — no appealable decision exists.
• Defensive value: This proceeding hands you nothing and costs you nothing. Unified did not invalidate a single claim and never got to institution, so there is no adjudicated invalidity to cite and no estoppel binding Unified under § 315(e)(2). What it does tell you is that a sophisticated defensive aggregator bought a settlement license rather than pressing the challenge — a data point that Unified's institution record was not viewed as a slam dunk, but also that Honeyman was willing to take money and walk away. The patents-in-campaign changed hands from Plano to Honeyman in March 2019, and the agreement means the current NPE has already demonstrated settlement receptivity.

IPR2017-01721 — State Farm Mutual Automobile Insurance Co. v. Plano Encryption Technologies, LLC

• Type: Inter Partes Review
• Filed: 2017-07-13 (Petition, Paper 1; Notice of Accord Filing Date 2017-08-08). Note a record discrepancy: the Board's termination judgment recites "State Farm filed its Petition on July 14, 2017 (Paper 1)," while the electronic file and the PTO Litigation Report show a 2017-07-13 filing date. Either way it was filed within one year of the 2016 district-court service, as the petition itself certified.
• Status: Verbatim: "JUDGMENT Termination of Proceeding 37 C.F.R. § 42.73." Google Patents records it as "PTAB case IPR2017-01721 filed (Settlement)."
• Judge panel: Joni Y. Chang, Annette R. Reimers, Michelle N. Wormmeester (same panel as the parallel IPR2017-01783 on companion patent 5,974,550)
• Petition grounds: Pre-AIA obviousness under 35 U.S.C. § 103, challenging claims 1, 2, 9, 10, 11, and 34 — the identical claim set Unified would later attack.
  • Ground 1 (§ 103) — claims 1, 9, 10, 11 obvious over David Aucsmith, "Tamper Resistant Software: An Implementation" (1996) in view of the knowledge of a person of ordinary skill. Petitioner argued Aucsmith uses nearly the same terminology, mapping "integrity verification kernel" onto the claimed "tamper resistant key module," with the key-pair generation step supplied by general cryptographic knowledge.
  • Ground 2 (§ 103) — claims 2 and 34 obvious over Aucsmith in further view of Chang (European Publication No. 0 686 906 A2), which allegedly taught distribution of a signed "software passport" between networked machines and supplied the "remote system / sending" limitations.
  • All grounds were pre-AIA; § 102 was not the thrust, § 112 was not pleaded.
• Institution decision: None. The Board found the proceeding "in the preliminary proceeding stage," with a Patent Owner Preliminary Response filed 2017-11-07/08 (Paper 6) and no institution decision entered.
• Final Written Decision: None. No paper adjudicates any claim on the merits.
• Settlement / termination: Joint Motion to Terminate filed 2018-01-05, resolving all disputes between the parties as to the '399 and '550 patents; the related E.D. Tex. district-court actions were dismissed. Termination judgment entered 2018-01-09; State Farm then requested refund of the post-institution fee (2018-01-09) and a Notice of Refund issued (2018-01-12). The Board noted that even if review had been instituted, State Farm would not participate further — no reply, no opposition to a motion to amend, no cross-examination of Plano's witnesses — which is precisely the posture that makes termination appropriate. Settlement terms are confidential (Ex. 1017).
• Appeal: None. (Note: the patent's litigation record lists a Federal Circuit case 16-2700, but that predates this 2017 IPR and is not an appeal from either PTAB proceeding. I have not confirmed its subject matter and won't characterize it.)
• Defensive value: Mixed, and mostly neutral-to-favorable. The petition was real, the art was serious, and the petitioner folded before institution—so there is still no estoppel and no invalidity ruling. The genuinely useful artifact here is the claimed-construction fight it surfaced over the word "including": Groupon argued in the related district-court campaign that the last-antecedent doctrine ties "including the generated private key and the encrypted predetermined data" to "the executable tamper resistant key module," so the module itself must contain the private key — not merely be built using one. Judge Gilstrap's August 2017 Markman order in the Alkami case (E.D. Tex.) endorsed that reading. Honeyman has since argued the opposite. That dispute, not any PTAB ruling, is your non-infringement engine.

Strategic summary

Claim-by-claim status: nothing is canceled, nothing is sustained, everything is untested. Claims 1, 2, 9, 10, 11, and 34 — the six claims both petitioners targeted — are the only claims that have ever been put at risk, and they emerged untouched because both cases settled before institution. Claims 3–8, 12–33, and 35–37 have never been challenged in any PTAB proceeding on this patent. So the honest answer to "is this patent narrowed?" is no. There is no canceled-claim list to hand a defendant, and there is no PTAB win to point to. What both petitioners hit is the same six-claim cluster; if you were drafting a petition today, the natural target set is the same claims — and if you wanted a margin of safety, claims 3–8 (which include the trusted-player-plays-decrypted-content and "specifically associated with the selected encrypted digital content" limitations) and claim 19's full conditional-use method are the untested ground where an institution record is thinnest.

Estoppel landscape: there is none, and that is the single most important defensive fact here. Under § 315(e)(2), estoppel attaches only to a petitioner that was a party to an IPR that resulted in a final written decision. Neither proceeding produced an FWD; in the 2017 case the Board went out of its way to say so, and in the 2020 case it said the termination order "does not constitute a final written decision pursuant to 35 U.S.C. § 318(a)." Consequently State Farm and Unified Patents are each free to raise again — and you are free to raise for the first time — Aucsmith, Chang/EP 0 686 906 A2, and any other ground. That means for a defendant currently being asserted against, the whole prior-art board is open: § 102 anticipation grounds were never pleaded at all, and the § 103 combinations were never tested against the intrinsic record. The only constraints you should plan around are discretionary, not estoppel-based: General Plastic factors if you file serially after your own earlier petition (not an issue for a first-time filer), and Fintiv/§ 315(b) if you've already been served and trial is advanced. And note the practical ceiling: the '399 patent shows an anticipated expiration of 2017-12-18 in the record, so any IPR would be against an expired patent (construed under Phillips), and the damages window has been closed for nearly nine years — meaning a petitioner's realistic motivation is back-damages exposure for pre-2017 conduct, not forward licensing.

Pattern signals. Two different petitioners, two different patent owners, eight years apart, the identical six-claim target set — that is a consistent signal that claims 1, 2, 9–11, and 34 are the commercially operative claims, and that "tamper resistant key module including the generated private key" is the crux. No petitioner has ever filed twice on this patent, so no same-petitioner serial-filing problem exists for anyone. A defensive aggregator is squarely in the chain: Unified Patents filed IPR2020-00213 against the current owner and then took a license, which means Honeyman Cipher has already been paid by the defensive-aggregator channel once. The patent owner has never pursued a PTAB appeal (nothing appealable exists) and has never defended an instituted trial. Ownership chain: Intel (1997) → Convera (2001) → H2 Partners (2011) → Plano Encryption Technologies (2015-04-03) → Honeyman Cipher Solutions (2019-03-08), with Plano/Liddle-associated entities driving the 2015–2016 campaigns and Honeyman driving the 2019–2020 campaigns.

Recommended next steps

1. Do not rely on PTAB invalidity to dispose of this patent. There is no FWD to cite, no canceled claim, and no estoppel. If your invalidity case is a repackaged Aucsmith + Chang theory, know that it has been presented twice and never once been evaluated by the Board, so its strength is genuinely unknown rather than "rejected."
2. Get the settlement documents if you can. The IPR2020-00213 Confidential Settlement and License Agreement (Ex. 1015) and the IPR2017-01721 settlement agreement (Ex. 1017) are sealed under 37 C.F.R. § 42.74(c) and 35 U.S.C. § 317(b), but they are obtainable "to any person on a showing of good cause." A license/exhaustion defense running through Unified's subscribers or through the 2018 State Farm resolution is a cheaper win than an IPR. That is the highest-value motion available to you.
3. Build your defense on the "including" construction. Pull Judge Gilstrap's August 2017 Markman order in the Alkami matter and the related E.D. Tex. claim construction orders, and the Groupon motion-to-dismiss papers attacking the infringement theory on the ground that the accused apps do not themselves include the private key. This is a § 112/§ 282(b) non-infringement fight, not a PTAB fight.
4. Check the calendar and the expiration date. If you were served more than one year ago, § 315(b) bars an IPR petition outright — confirm your service date first. Separately, the record shows expiration on 2017-12-18; scope your exposure to the § 286 six-year lookback from the operative complaint, and cap it at the expiration date.
5. If you do file, target the same six claims and bring § 102. Both prior petitions were § 103-only. Anticipation, and the Honeyman v. Groupon claim-construction record, are the two untested levers on this patent.

Primary sources:

• IPR2020-00213, Decision — Settlement Prior to Institution of Trial, 37 C.F.R. § 42.74 (2020-03-25), Paper 8: docketalarm.com PDF · Unified Patents case portal
• IPR2017-01721, Judgment — Termination of Proceeding, 37 C.F.R. § 42.73 (2018-01-09): docketalarm.com PDF (consolidated with IPR2017-01783) · document index at Patexia · Unified Patents case portal
• Unified Patents, "Honeyman Cipher settles with Unified Patents" (2020-03-23): unifiedpatents.com
• Underlying patent page: US5991399A on Google Patents

Confidence caveats: I confirmed both case numbers, both petitioners, both patent owners, both panels, the challenged claim sets, and both termination events from primary Board documents. I did not locate the specific prior-art references cited in the Unified Patents IPR2020-00213 petition (no public adjudication exists), and the exact commercial terms of both settlements are confidential — I have not guessed at them. The IPR2017-01721 petition filing-date discrepancy (2017-07-13 vs. 2017-07-14) is noted rather than resolved.

I'll start by pulling the assignment and litigation record from multiple sources.

Good initial hits. Let me dig for reel/frame data, correspondents, and the litigation/assertion history.

Ownership & Assignment Analysis — US 5,991,399

Data-retrieval caveat up front: I was able to reconstruct the chain of assignees and the litigation record, but I could not retrieve USPTO reel/frame numbers or the recorded correspondent (recording attorney/agent) for any link. The USPTO Assignment Center (https://assignmentcenter.uspto.gov — mirrored at https://assignment.uspto.gov/patent/index.html) was not queryable in this session; web searches surfaced the Google Patents legal-events list and, for one link, the full assignment deed text filed as a PTAB exhibit. Where a value is unknown I say so rather than guess. The two data points the brief cares most about — reel/frame and recording correspondent — should be pulled directly from Assignment Center by patent number 5991399 to close the gap.

Inventors

All four appear together as assignors on the original "ASSIGNMENT OF ASSIGNORS INTEREST" recorded 1997-12-18 to Intel Corporation, the same day the application was filed. That is the standard employee-invention assignment pattern for a corporate filing; there is no evidence of outside inventors or joint development.

Pattern notes:

• I could not determine inventor departure dates or confirm whether any left Intel within 12 months of filing. That check (often a pre-fire-sale tell) is unclear on the available record.
• One oddity worth recording: the 2015 court AO-120 forms in National Bank v. Plano Encryption Technologies (6:15-cv-00249, W.D. Tex.) list the "holder" of the '399 patent as Richard L. Maliszewski — a named inventor, not an assignee. By that date title had already moved to Plano Encryption Technologies (assignment of 2015-04-03). This is almost certainly stale PTO assignment data auto-populating the court form, not a retained ownership interest by the inventor. Flagging it because it is exactly the kind of artifact that gets mis-read as an inventor-owned patent.

Original assignee

Intel Corporation (Santa Clara, CA). Application filed 1997-12-18; patent issued 1999-11-23.

• Line of business: semiconductors / PC platform. The patent's own framing is PC-centric — the specification describes the "open and accessible" PC as "a fundamentally insecure computing platform" and the trusted player as a DVD or CD-ROM player on that platform. This is consistent with Intel's platform business and its contemporaneous work in the DVD/content-protection arena.
• Did Intel ship a product embodying the claims? Not established on the records I retrieved. The claims are method/apparatus claims to generating a tamper-resistant key module and distributing it to a remote trusted player — a server-side key-compilation and delivery technique. I cannot point to a shipping Intel product that practices it, and I will not assert one without evidence.
• Current status of Intel: operating company (public, NASDAQ: INTC). Intel was not in bankruptcy at the time of the 2001 transfer; the Intel→Convera transfer is a corporate divestiture, not a distress sale.

Assignment timeline

Every recorded link I could identify, oldest first. Reel/frame and correspondent are marked not retrieved where that is the honest state of the data.

1997-12-18 (executed) / recorded 1997-12-18 — Reel not retrieved

• Conveyance: Assignment (Assignment of Assignors' Interest)
• Assignor: Gary L. Graunke; John Carbajal; Richard L. Maliszewski; Carlos V. Rozas
• Assignee: Intel Corporation
• Correspondent: not retrieved
• Context: Initial employee-inventor assignment to employer, recorded same day the application was filed.

2001-07-23 — Reel not retrieved

• Conveyance: Assignment
• Assignor: Intel Corporation
• Assignee: Convera Corporation
• Correspondent: not retrieved
• Context: Divestiture — Intel's interactive-media/content-protection assets were folded into Convera (the entity formed around Excalibur Technologies and Intel's interactive media group). Not a distress transfer.

2011-01-20 — Reel not retrieved

• Conveyance: Assignment
• Assignor: Convera Corporation
• Assignee: H2 Partners LLC (New Canaan, CT)
• Correspondent: not retrieved
• Context: Portfolio disposition out of a wound-down/changed operating company into a bare holding vehicle with no products. Convera's status at transfer is unclear on my record — this is a fire-sale-vs-ordinary-disposition question that needs SEC 10-K/8-K confirmation for Convera's 2010–2011 fiscal years.

2015-04-03 (executed; see date note) / recorded 2015 — Reel not retrieved

• Conveyance: Assignment (titled "Assignment of Patent Rights")
• Assignor: H2 Partners LLC, "with a location at New Canaan, CT"
• Assignee: Plano Encryption Technologies, LLC, "having a primary place of business at Plano, TX"
• Correspondent: not retrieved (the deed text I recovered does not carry the USPTO recording correspondent block)
• Context: Transfer to an asserting vehicle. The deed conveys not just the patents but expressly "(i) causes of action and enforcement rights … all rights to pursue damages, injunctive relief and other remedies for past, present and future infringement." It covered a portfolio: US 5,991,399; 5,974,550; 6,041,122; 6,529,603 plus applications 09/441,409; 09/603,079; 09/709,901. The deed text reads "executed at 11:30am …18,2015," with the day partially garbled in the source; Google Patents' legal events list 2015-04-03. Date is approximate — verify at Assignment Center.
• Corporate structure (from Plano's own IPR filings): Plano Encryption Technologies, LLC is a Texas LLC and is a wholly owned subsidiary of Koba Holdings, LLC (Delaware). Both are named real parties-in-interest.

2019-03-08 — Reel not retrieved

• Conveyance: Assignment
• Assignor: Plano Encryption Technologies, LLC
• Assignee: Honeyman Cipher Solutions LLC
• Correspondent: not retrieved
• Context: Transfer to a fresh asserting vehicle immediately ahead of a second litigation campaign. Unified Patents characterizes Honeyman Cipher as a "Brad Liddle entity and well-known NPE."

Other recorded legal events (not assignments):

• 2017-12-18 — Anticipated expiration. The patent term ended here, yet Honeyman asserted it in 2019–2020 (past-damages campaign).
• PTAB: IPR2017-01721 (filed, terminated by settlement) and IPR2020-00213 (filed 2019-12-13 by Unified Patents against the '399 patent, joint motion to terminate on settlement 2020-03-20).

Note on §1 above: Assignment Center does have records for this patent — at minimum the five links above. I am not stopping at the timeline section for that reason; the gap is specific to reel/frame and correspondent, not to the existence of records.

Timeline diagram

timeline
    title Ownership of US 5991399
    1997 : Filed by Intel Corp
         : Inventors assign to Intel
    1999 : Patent issued
    2001 : Assigned to Convera Corp
    2011 : Assigned to H2 Partners LLC
    2015 : Assigned to Plano Encryption Technologies
         : First infringement suits filed
    2017 : Patent term expires
    2019 : Assigned to Honeyman Cipher Solutions
         : New litigation campaign begins
    2020 : Unified Patents IPR settled

NPE / troll-pattern signals

1. Shell-entity transfer — PRESENT.
Title moved out of Intel to Convera, then to H2 Partners LLC (New Canaan, CT — a bare holding vehicle with no products identified), then to Plano Encryption Technologies, LLC, a Texas LLC whose own IPR filings describe it only as a patent owner, held as a wholly owned subsidiary of Koba Holdings, LLC (Delaware), and then to Honeyman Cipher Solutions LLC. The 2015 deed is the tell: it conveys the patents plus all past/present/future infringement causes of action — the drafting of a vehicle built to sue. No products in commerce were identified for H2, Plano, Koba, or Honeyman on any source I retrieved. Reel/frame verification pending.

2. Known asserter in the chain — PRESENT.
Both Plano Encryption Technologies and Honeyman Cipher Solutions are publicly identified as NPEs by Unified Patents. Unified's December 2019 post calls Honeyman "a Brad Liddle entity and well-known NPE" and Plano "another NPE." Neither name appears on the classic enumerated list (Acacia, Marathon, IV, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Round Rock, etc.), but the brief also asks for "any entity surfaced by Unified Patents or RPX as a high-frequency plaintiff," and these qualify: Plano filed/defended 20+ actions on this family in 2015–2016 (banks, State Farm, Best Buy, Etsy, Shutterfly, Match.com, Groupon, J.C. Penney, Q2 Holdings, Alkami); Honeyman filed a second wave in 2019–2020 (Snap, Groupon, LogMeIn, Evernote, Adidas, Asics Digital, Asana, PayPal, Slack).

3. Repeat correspondent across the chain — UNCLEAR (data not retrieved).
I could not obtain the USPTO-recorded correspondent for any assignment link, so I cannot call recurrence. Two adjacent data points, offered expressly as not proof of the assignment-recording correspondent:

• PTAB records for IPR2020-00213 list David M. Quinlan, P.O. Box 8088, 22 Monterey Drive, Princeton, NJ 08543 (quinlapc@msn.com) as counsel for Honeyman Cipher.
• Plano's mandatory notice in the sibling IPR lists Bradley D. Liddle as counsel for Plano Encryption Technologies — i.e., the NPE principal appearing as its own attorney of record.

These are litigation/PTAB counsel of record, not recorded assignment correspondents. Action item: pull the correspondent block for all five links at Assignment Center and check whether a single recording attorney/firm repeats across H2 → Plano → Honeyman. That recurrence, if it exists, is the strongest remaining signal and it is currently unverified.

4. Cascading transfers (<24 months through chained LLCs) — NOT PRESENT.
The links are widely spaced: Intel→Convera 2001; Convera→H2 2011 (~10 yrs); H2→Plano 2015 (~4 yrs); Plano→Honeyman 2019 (~4 yrs). No sub-24-month cascade. The portfolio was clearly assembled and re-packaged over ~18 years rather than flipped.

5. Pre-litigation transfer — PRESENT, TWICE.

• H2 → Plano assigned 2015-04-03; first Plano suit 2015-06-29 (Plano Encryption Technologies v. Citizens National Bank, 2:15-cv-01168, E.D. Tex.) — ~3 months.
• Plano → Honeyman assigned 2019-03-08; first Honeyman suit 2019-06-05 (Honeyman Cipher Solutions v. Groupon, 1:19-cv-03754, N.D. Ill.) — ~3 months.

Both fall inside the 6-month window and match the classic clean-standing/venue-setup pattern.

6. Bankruptcy fire-sale — NOT PRESENT.
No Chapter 7/11 for Intel or Convera was identified. Intel was and is an operating company. Whether Convera's 2011 disposal was a wind-down sale is unclear (requires Convera 2010–2011 SEC filings).

7. Privateering — NOT PRESENT as to Intel; UNCLEAR as to Convera.
The 2001 Intel→Convera transfer reads as a business divestiture (Intel's interactive-media group rolled into Convera), not as Intel arming an NPE against competitors. Whether Convera's 2011 transfer to H2 Partners was a genuine disposition or an arranged hand-off is not established on my record.

8. Defensive aggregator — NOT PRESENT.
The chain terminates at Honeyman Cipher Solutions LLC, an asserting entity — the opposite of neutralization. Unified Patents challenged the patent (IPR2020-00213) but settled and terminated on 2020-03-20 rather than acquiring it; IPR2017-01721 likewise settled. No RPX / AST / LOT / OIN acquisition appears anywhere in the chain.

Verdict

NPE — high confidence.

Three strong signals converge. (a) Shell-entity transfer: the 2015 deed from H2 Partners LLC to Plano Encryption Technologies, LLC conveys the patents together with all past, present and future infringement causes of action, and title then passes to Honeyman Cipher Solutions LLC — three successive non-operating holders, with Plano itself disclosed in its IPR papers as a wholly owned subsidiary of Koba Holdings, LLC. (b) Known asserter: Unified Patents publicly labels Honeyman "a Brad Liddle entity and well-known NPE" and Plano "another NPE," with 20+ suits filed on this family. (c) Pre-litigation transfer, twice: H2→Plano on 2015-04-03 followed by suit on 2015-06-29, and Plano→Honeyman on 2019-03-08 followed by suit on 2019-06-05 — each roughly three months, well inside the six-month window.

The single unresolved link that could move this from "high" toward a forensic-grade record is signal #3 (repeat recording correspondent). The chain's assignee names are confirmed; the reel/frame and correspondent fields are not — those must be read off Assignment Center for patent 5991399 before the correspondent-recurrence call can be made either way.

Verification links:

• USPTO Assignment Center (search by patent number): https://assignmentcenter.uspto.gov
• USPTO Assignment Search (legacy interface): https://assignment.uspto.gov/patent/index.html
• Google Patents legal events for this patent: https://patents.google.com/patent/US5991399/en
• Unified Patents on the IPR filing (2019-12-13): https://www.unifiedpatents.com/insights?offset=1576612995741
• Unified Patents on the settlement (2020-03-23): https://www.unifiedpatents.com/insights/2020/3/23/honeyman-cipher-settles-with-unified-patents

I'll search for the USPTO record and the prior-art citations listed for US 5,991,399.

Let me pull the front-page "References Cited" section specifically, since the fetched patent text didn't include it.

Let me check the USPTO full-text record and the original front page of the patent.

Search performed and key finding

I searched for the USPTO/Google Patents record of US 5,991,399 and specifically for its backward citation list (the "References Cited" / "Patent Citations" front-page section that lists prior art the examiner or applicant cited).

Finding you should note up front: the authoritative full text of US 5,991,399 (Google Patents rendering of the USPTO record, fetched 2026‑09‑29) that you supplied does not contain a "Patent Citations" / "References Cited" section at all. The only citation block in that record is:

Cited By (422) — i.e., forward citations (later patents/third parties citing '399), not prior art for '399.

So per your own precedence rule (prefer the authoritative patent text; apply it literally), the record in front of us does not list any cited U.S. patent documents for 5,991,399. I therefore cannot honestly enumerate "each patent citation for 5991399" — there is no such list in the authoritative source. I'm flagging this as a contradiction with the premise of the request rather than inventing a citation list.

Below I give everything the record does support, and separate it clearly from what it does not.

A. What the authoritative record for '399 contains

B. The only references "cited" inside the '399 text (specification, not front page)

These are the two commonly-assigned, co-pending applications that the '399 specification expressly identifies and incorporates by reference. In substance they are the closest thing the patent has to internal citations:

§ 102(b)-type analysis (non-publication) — relevance to scope only, not anticipation:

• Both are drawn to the tamper-resistant compiler / IVK / bilateral-authentication machinery, which is the enabling substrate for '399's "executable tamper resistant key module."
• Ser. No. 08/662,679 → U.S. 5,892,899 (filed 1996‑06‑13, before '399's 1997‑12‑18 priority) is potentially § 102(e) prior art on its face, subject to whether it shares an inventive entity/common inventor with '399 (inventors here: Gary L. Graunke; John Carbajal; Richard L. Maliszewski; Carlos V. Rozas). I could not verify U.S. 5,892,899's inventorship, so I will not assert it anticipates anything. Its disclosure, as characterized by '399 itself, covers the anti-tamper/IVK/bilateral-authentication apparatus — not the '399-specific sequence of "generate fresh asymmetric pair → encrypt data with the public key → build a module containing the private key + encrypted data." It is more plausibly § 103 fodder / antecedent art than a § 102 anticipation of claims 1, 19, 20, 34–37.
• Ser. No. 08/924,740 (filed 1997‑09‑05) is not § 102(b) art (unpublished, if it published at all, it must be § 102(e) via a granted patent) and, like the first, does not appear to disclose the '399 conditional-use-key-module sequence.

Honest conclusion for Section B: Neither incorporated application is a § 102 anticipation of any specific claim of '399 on the face of the '399 disclosure. They are the architectural backdrop '399 builds on.

C. Trap to avoid — the "Patent Citations (6)" snippet is not '399's list

My searches surfaced a snippet reading "Patent Citations (6)" immediately adjacent to "US5991399A." That block belongs to US 6,550,011 B1 ("Media content protection utilizing public key cryptography," Hewlett‑Packard), which cites '399 as one of its six citations. It is not a citation list of '399. Similarly:

• US 6,438,235 B2 (Hewlett‑Packard) — cites '399 (forward).
• US 6,567,917 B1 (Cisco, "Method and system for providing tamper-resistant executable software") — cites '399 (forward).
• US 2003/0097579 A1 (England / Microsoft, "Manifest-based trusted agent management…") — cites '399 (forward).

All of these post-date '399's 1997‑12‑18 priority and are therefore not prior art against '399.

D. Evidence of what the field treats as the relevant technical content of '399

External search reports about later cases cite '399 as prior art — useful for understanding the examined scope, but again these are citations OF '399, not BY '399:

(Note: the OCR of the JP report renders the lead inventor as "MALISZEKKI RICHARD L ET AL." Per your literal-interpretation rule I have not auto-corrected it, but the authoritative '399 text gives Richard L. Maliszewski.)

The [X] treatment by the EP examiner in EP 1 370 997 is the strongest external signal that a single reference was viewed as capable of reading on '399's broadest subject matter. It bears on claim 1-type scope, but I do not have EP 1 370 997's text in hand, so I will not manufacture a limitation-by-limitation § 102 mapping.

E. § 102 anticipation mapping — what I can and cannot say

A § 102 anticipation requires one reference disclosing every limitation of the claim, arranged as in the claim. Using the claim paraphrases from the previously generated section:

Bottom line: From the authoritative '399 record alone, I cannot attribute any anticipating § 102 reference to any specific claim of 5,991,399, because the record supplies no backward patent citations and I could not retrieve the USPTO PatentCenter "References Cited" list or the international search report for the '399 PCT sibling (below) within my tool budget.

F. What would complete this analysis (records I could not retrieve)

1. USPTO PatentCenter "References Cited" for application US 08/993,597 (patent 5,991,399) — the official front-page citation list. My searches did not surface this list; the Google Patents rendering you supplied omits it.
2. International Search Report for PCT/US1998/026415 (the '399 PCT sibling, published as WO 1999/031842 A1) — the ISR "documents considered to be relevant" is the best proxy for the art the '399 family was examined against.
3. The full text of the two incorporated-by-reference applications (Ser. No. 08/662,679 → U.S. 5,892,899; Ser. No. 08/924,740) including their inventorship, to test the § 102(e) question in Section B.
4. IPR2017‑01721 and IPR2020‑00213 petition papers (Unified Patents v. Honeyman Cipher Solutions) — these would name the specific prior-art patents the petitioner asserted against '399's claims 1, 2, 9–11, 34 (the claims identified as asserted in the Honeyman v. Snap complaint per the prior section). That is likely the single richest source of a concrete § 102/§ 103 mapping.

G. Caveats (explicit)

1. No fabricated citations. I did not enumerate a "Patent Citations" list because the authoritative record contains none; producing one from memory would violate your literal-interpretation rule.
2. Forward ≠ backward. The 422 "Cited By" entries are not prior art to '399.
3. Snippet misattribution flagged. "Patent Citations (6)" in search results belongs to US 6,550,011 B1, not to '399.
4. Section B § 102(e) status is conditional on inventorship of U.S. 5,892,899, which I could not verify.
5. The two incorporated applications are described as "commonly assigned to the same entity as the present invention" (Intel) in the '399 text — so common-ownership considerations under pre-AIA § 103(c) would bear on any obviousness combination, though '399's § 102 novelty questions are a separate matter.
6. My searches hit the tool-step limit before I could pull the USPTO PatentCenter citation list directly; the absence of a backward-citation section in the supplied authoritative text is my primary basis above, and I state it as a finding about that record, not as a proven fact about the certificate of the printed patent.

Sources relied on: the authoritative US5991399A text/claims you supplied (Google Patents, fetched 2026‑09‑29); Google Patents pages for US6550011B1, US6438235B2, US6567917B1, US2003/0097579A1 (forward citations of '399); EPO search-report data for EP 1 370 997 A4 and EP 1 480 100 A1; PCT/IB03/03696 international search report as published in JP 2005‑536797 A; KR 10‑0971854 B1 search-report citation list (forward citation of '399).

I'll ground this analysis in the actual prior-art record. The fetched page text includes a large forward-citation table ("Cited By") but the backward-citation ("References Cited"/prior art) list did not come through in the supplied text, so let me try to retrieve it before analyzing.

The supplied page text includes the forward-citation table but not the front-page "References Cited" list. Let me try to retrieve the actual prior-art references of record for this patent.

The page's back-citation list didn't come through, so let me verify the specific prior-art references that are actually relevant to this subject matter (tamper-resistant software, trusted-player key distribution, and certificate/manifest-based integrity).

Obviousness Analysis — U.S. Patent 5,991,399 (35 U.S.C. § 103, pre‑AIA)

Analysis date: April 26, 2026 (note: the authoritative page text I was given carries a fetch timestamp of 2026‑09‑29; I flag the two dates as inconsistent rather than reconcile them)

1. Provenance / evidence limitation — read this first

The supplied page content does not contain the "References Cited" (backward prior-art) section. What the fetched text actually contains is:

• the "Prior art keywords" field (key, tamper resistant, executable, key module, integrity),
• the specification's own "Description of Related Art" and "Definitions,"
• a "Cited By (422)" table — i.e., forward citations, and
• the claims (37), classifications, family, and assignment/litigation data.

I attempted to retrieve the examiner-cited front page ("References Cited") for US5991399 with two targeted searches. Both returned only (a) the same Google Patents page without a Citations block, (b) third-party search reports in which '399 is itself cited as prior art (e.g., the EPO ISR for WO 2004/019189, which lists US 5 991 399 A … (Y) … 1,2,4–7), and (c) unrelated "References Cited" tables from other patents.

Consequence: I cannot do a claim chart against "the references cited by the examiner," because that list is not in the record I was given, and my retrieval attempts failed. Rather than fabricate pin cites, I build the analysis from three evidence tiers, each labeled:

I also flag one internal contradiction in the record and one conflicting attribution in §7 below, and one statutory hurdle that materially weakens the strongest combination in §5.

2. Governing law and the correct analytic frame

• Pre-AIA § 103 governs (application filed 1997‑12‑18; AIA § 3 applies to applications filed on/after 2013‑03‑16). So the analysis is under Graham v. John Deere Co., 383 U.S. 1 (1966), as flexibly applied by KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007).
• Critical date: 1997‑12‑18 for § 102(b)/(a) purposes; the invention date could be earlier but is not established on this record.
• Combination standard: it is not required that references be physically combinable into one structure; the test is whether the claimed subject matter as a whole would have been obvious from the references considered together. In re Keller, 642 F.2d 413 (CCPA 1981).
• Applicant's own admissions count as prior art for § 103. Statements in the '399 specification characterizing the state of the art and its deficiencies are binding admissions. Riverwood Int'l Corp. v. R.A. Jones & Co., 324 F.3d 1346 (Fed. Cir. 2003); In re Nomiya, 509 F.2d 566 (CCPA 1975). This is unusually important here because the '399 background is essentially a roadmap of the invention.
• "Obvious to try" applies where the prior art gives a finite number of identified, predictable solutions. KSR; In re Merck & Co., 800 F.2d 1091 (Fed. Cir. 1986). Tamper-resistant compilation applied to a key-bearing executable is such a case (one design choice among a small set of known key-hiding techniques).
• Teaching away requires a reference that criticizes, discredits, or otherwise discourages the solution — mere disclosure of alternatives is insufficient. In re Gurley, 27 F.3d 551 (Fed. Cir. 1994); In re Fritch, 972 F.2d 1260 (Fed. Cir. 1992). See §6 for the one place this genuinely bites.

Practical note on stakes: the patent expired (anticipated expiration 2017‑12‑18 per the Google Patents record). Obviousness therefore has little prospective effect; it would matter only to the historical damages window (roughly 2011–2017) in the Honeyman/Plano campaigns, and to any re-examination-style challenge. I found no 2026 activity and no 2026 CAFC docket, consistent with the earlier generated section.

3. Person of ordinary skill in the art (POSITA)

For the 1997‑12‑18 critical date, a POSITA would be a software engineer/systems security practitioner with a B.S. in CS/EE and ~2–4 years' experience, or equivalent, having working knowledge of: (i) public-key and symmetric cryptography and hybrid key transport ("wrap a session key under a public key"); (ii) digital signatures, one-way hashes, and certificates (the '399 background itself recites DSA, DSS, NIST, hashing, and root keys — T1); (iii) PC software engineering including loaders, plug-in architectures, and object-code linking; and (iv) the then-emerging tamper-resistant/obfuscating-code literature. The claim language itself assumes this level ("key compiler," "Monkey components"/Montgomery arithmetic is recited in the spec, not the claims).

4. Element-by-element mapping of the independent claims

The table maps claim limitations to the strongest available evidence tier. "T1/T4" means the element is conceded or described as known in the specification itself and/or is textbook general knowledge.

4.1 Claim 1 (broadest — stops at "building")

Claim 1 is the most exposed claim. It requires no transmission, no verification, and no decryption. Every limitation is either admitted in the specification or taught by the two applications the specification incorporates by reference.

4.2 Claim 19, 20 (end-to-end loop), 27–29, 31 (CRM/apparatus/system)

4.3 Claims 34–37 (minimal build / receive family)

Claim 34 recites only building the executable tamper-resistant module (private key + ciphertext) and sending it — it omits key generation and encryption entirely. Claims 35/37 recite the mirror-image receive/execute/decrypt. These are the most vulnerable claims in the patent: they are effectively "ship a tamper-resistant key-bearing executable" and "run it." Any § 103 combination that reaches claim 1 or claim 19 also reaches 34–37 a fortiori, and a single reference teaching a tamper-resistant executable containing a decrypt key would be § 102 art for claim 34. This is consistent with why the earlier-generated summary notes the Snap complaint asserted "at least exemplary claims 1–2, 9–11, 34" — the assertor picked the broad end, which is also the obviousness-exposed end.

5. The combinations

I give four: one primary (the whole-claim-set attack) and three that fill specific gaps, plus a statutory obstacle the challenger must clear.

Combination 1 — "Tamper-resistant executable key carrier + standard public-key (hybrid) key transport" (PRIMARY)

References:

• [A] US 5,892,899 (Intel; Ser. No. 08/662,679, filed 1996‑06‑13) and [B] Ser. No. 08/924,740 (filed 1997‑09‑05) — both incorporated by reference into '399 itself. Teach: a tamper-resistant compiler that takes a prepared code module and emits a self-modifying, self-decrypting image; integrity verification kernels; bilateral authentication between cooperating modules; validation of object code on disk and in memory; secure linkage via address validation. (T2)
• [C] The admitted art of '399's own background — a DVD player with an embedded key that decrypts content and plays it in real time; key distribution under a public/private pair; certificates and a root key. (T1)
• [D] General knowledge — asymmetric encryption of a session/content key under the recipient's public key, and per-title content keys in conditional-access/DVD systems. (T4)

How the combination reads on the claims. [C] supplies the content-decryption architecture and the conceded crypto primitives; [D] supplies the "generate a fresh pair, encrypt the payload under the public key" step (which is literally described in '399's background as the standard use of public-key encryption); [A]/[B] supply the "executable tamper resistant module" that carries the key and self-checks, plus the IVK/bilateral-authentication machinery that performs the program- and module-integrity checks recited in claims 3, 4, 19, 20, 31, 35 and 37. The mapping of the key-compiler idea itself is admitted in '399's Definitions section.

Motivation to combine (KSR-compliant). The references, taken together, address the identical problem stated in '399's own background: an open, insecure PC whose software "could be 'hacked' and the key obtained." '399 expressly frames the deficiency of the prior art as (i) a global key usable on all content and (ii) a key that is pre-loaded. The two incorporation-by-reference applications exist precisely to make software resistant to observation and modification. A POSITA reading them side by side has: a stated problem (extract the key from a hacked player), a known class of solutions (obfuscate/encrypt the code so it only executes if unaltered), and a known payload (the content key). Applying a tamper-resistant compiler to the very module that carries a session key is the predictable use of a known technique for its known purpose, with a reasonable expectation of success — the classic KSR "improvement in one recognized technique" formulation. Secondary indicia: market pressure from DVD/Internet piracy, expressly recited in the background, supplies the design incentive.

Claims reached: 1, 2, 3, 4, 5, 6, 7, 9, 12, 13, 14, 15, 16, 17, 19, 20, 27, 28, 29, 31, 34, 35, 36, 37. Dependent claims 9–14 (IVK generation, manifest parser generator, key compiler, tamper-resistant compiler) are directly met by [A]/[B], whose whole subject matter is IVK generation and bilateral authentication — so the usual "dependent claim adds implementation detail" defense largely evaporates.

Combination 2 — "Authenticate-the-receiving-node-then-decrypt" (fills the conditional in "conditional use")

Reference: US 5,633,932 — Apparatus and method for preventing disclosure through user-authentication at a printing node, Davis et al., issued 1997‑05‑27 (verified: https://uspto.report/patent/grant/[5633932](/patent/5633932) ; https://patents.searchlight.law/doc/US5633932). Discloses: a sending node encrypts a document under the receiving node's public key; the receiving node stores the ciphertext, withholds output until the requesting party is authenticated, and then decrypts; the header carries control information (e.g., "print-only" tags) that restricts downstream use; node certificates and locally issued verification certificates establish the node's characteristics.

Combined with Combination 1's tamper-resistant loader/IVK and with the admitted crypto of [C]/[D].

Motivation. Davis supplies the missing concept for claims 3/4/19/20/25/32/35/37: decryption is contingent on a successful authenticity check on the consuming side, plus usage-control metadata carried with the protected object. '399 supplies the code-integrity dimension Davis lacks (Davis authenticates the recipient, not the executing program). The combination is motivated because both are "vessel"-protection systems in the same technical field (protect content until the consuming node proves it is compliant), and the addition of code-integrity checking to recipient-authentication is an enhancement of one technique by another, with predictable results. No teaching away: Davis teaches that node characteristics may be certified and that output is conditioned on them — the logical next step is to certify the player program rather than (or in addition to) the person.

Claims reached: 3, 4, 5, 18 (authorization/financial information carried in the request), 20, 25, 26, 31, 32, 33, 35, 37.

Combination 3 — "Per-title key + server-side transaction" (fills claims 7, 8, 15–18, 21–24)

References: [C] admitted conditional-access art plus [D] general knowledge of title-key/content-key schemes in which the decryption key is specific to the item and delivered under a license or session key after a purchase. '399's own background contrasts a "global" key with the need for per-content keys, which is an admission that the desirability of per-title keys was known ('399's stated advantage is that "each key is valid only for selected digital content").

Motivation. Once the specification concedes that the problem with the prior art is a global key, the substitution of a per-title key and a per-title request is the direct, predictable remedy the specification itself identifies — an admission-driven obviousness case (Riverwood). Claims 16–18 and 23–24 (create content, encrypt with plural symmetric keys, store keys, request them with content identifier, client identifier, and payment information) add only conventional electronic-commerce plumbing: billing identifiers and credit-card fields in a request message.

Claims reached: 5, 7, 8, 15, 16, 17, 18, 21, 22, 23, 24, 32.

Combination 4 — "Obfuscating/tamper-resistant code as a known technique" (independent § 102(b)/§ 103 support)

Reference candidates (T3/T4, in descending confidence):

1. Aucsmith, "Tamper Resistant Software: An Implementation," Proc. First Int'l Workshop on Information Hiding, Cambridge, 1996 — the published companion to the Intel patent family. Moderate-high confidence in existence/date; not verified in this session.
2. F. Cohen, "Operating System Protection Through Program Evolution," Computers & Security 12(6), 1993 — self-modifying/evolving code as a protection mechanism. Moderate confidence.
3. Menezes, van Oorschot & Vanstone, Handbook of Applied Cryptography (1996) and Schneier, Applied Cryptography, 2d ed. (1996) — for RSA/DSA, one-way hashes, key wrapping, and certificate chains. High confidence as general knowledge.

Why this matters more than it looks: printed publications are § 102(b)/(a) art without regard to common ownership. This is the escape hatch from the statutory obstacle in the next subsection. The publications also supply the reason a POSITA would apply obfuscation to a key-bearing module: hiding secrets in executable code was a recognized discipline by 1997.

5.1 The statutory obstacle the challenger must plan around (important)

Under pre-AIA 35 U.S.C. § 103(c), subject matter that qualifies as prior art only under § 102(e), (f), or (g) cannot be used in a § 103 rejection where, at the time the invention was made, that subject matter and the claimed invention were commonly owned or subject to an obligation of assignment to the same person. '399 is an Intel application; US 5,892,899 and Ser. No. 08/924,740 are Intel filings. If those two applications are § 102(e) art only (US patent granted on a pre-critical-date U.S. filing — US 5,892,899 issued 1999‑04‑06 on a 1996‑06‑13 filing, so § 102(e) by its terms applies, provided the inventive entity is "another," which appears to be the case and should be verified), then § 103(c) disqualifies them from the § 103 combination.

That does not kill the attack; it redirects it:

1. Prior-art admissions. The '399 specification's own descriptions of IVKs, tamper-resistant compilation, bilateral authentication, and the key compiler are admissions usable as evidence of the state of the art (Riverwood, Nomiya), independent of § 103(c).
2. Printed publications. The Aucsmith 1996 paper, Cohen 1993, and the 1996 crypto treatises are § 102(b) art and are not subject to § 103(c) because they are not § 102(e)/(f)/(g) art.
3. Unrelated third-party art. The Davis reference (T3) and the general conditional-access/DVD art carry no common-ownership problem.

A competent challenger would therefore lead with published art, using the Intel applications and the specification admissions as corroboration and as the motivation-to-combine narrative — not as the load-bearing § 102(e) references.

6. Where the obviousness attack is weakest (the honest counter-case)

1. The "teaching away" argument — and why it is weaker than it first appears. Classical key-management doctrine (Diffie/Hellman lineage; restated in every 1996 treatise) holds that a private key is never transmitted or shared. '399 does the opposite: it ships a private key to the client inside an executable. This is the patent's real inventive kernel. The rebuttal is strong, though: the doctrine protects long-term identity keys. '399 generates an ephemeral, single-use, content-specific key pair whose only function is to unwrap one payload for one transaction, and it is generated by the same party that sends it. Diffie-Hellman doctrine is not even addressed to, let alone critical of, that scenario — and under In re Gurley a mere alternative disclosure is not a teaching away. Net: a genuine argument, but not a likely winner on its own.
2. Claim 10's specific pipeline. "Accessing an asymmetric public key… associated with a manifest of the program signed by an asymmetric private key… producing IVK code… combining manifest parser generator code and the IVK code" is the most implementation-specific limitation. If the published art (rather than the commonly-owned applications) were the only § 103(c)-eligible evidence, a challenger could struggle to show that the specific manifest-parser-GEN + key-compiler integration was suggested rather than merely obvious to try. This is the single most defensible claim in the patent on obviousness grounds, together with claims 11–14 insofar as they track it. It is also narrow, and '399's own assertions went to claims 1–2, 9–11, 34 — suggesting even the patentee treated the broad end as the commercial center of gravity.
3. Secondary considerations. I found no evidence of the classic nexus-bearing indicia (unexpected results, industry skepticism, licensing due to the merits, long-felt need) in the material available. In an NPE posture, licensing revenue is unlikely to support nexus. I state this as an absence of evidence, not as evidence of absence.
4. Predictability caveat under In re O'Farrell. "Obvious to try" must not be stretched to a "try every permutation" rationale; the tamper-resistant-compiler route must be shown to have been identified as a solution, not merely a possibility. The 1996 Aucsmith publication and the incorporated Intel applications serve exactly that function — which is why Combination 4's publication evidence is the linchpin, not an afterthought.

7. Record conflicts and unverified items I am flagging, not resolving

1. Forward citations are not prior art. The page's "Cited By (422)" table lists some documents with pre‑1997 "priority dates" (e.g., US6546193B1 at 1997‑10‑28; US20030069070A1 at 1997‑05‑28). Google Patents' "priority date" column for a citing document does not establish that document as § 102 art against '399 — a document that cites '399 as prior art is presumptively later in the relevant chain, and its actual filing/§ 102(e) chain must be checked case by case. I therefore did not use forward citations as prior art.
2. Conflicting attribution for US 5,224,163. The EPO ISR in WO 2004/019189 lists it as US 5 224 163 A (KAUFMAN CHARLES W ET AL) 29 June 1993 and cites it as a "Y" reference; Google Patents/Justia-derived listings I retrieved show 5,224,163 | June 29, 1993 | Gasser et al. I have not auto-corrected either and I am not relying on this reference for any limitation. It is a retrieval lead only.
3. Inventor-name OCR variants. The third-party reports render the surname as NALISZEWSKI and MALISZEKKI; the authoritative patent text gives Richard L. Maliszewski. Recorded, not corrected.
4. Unverified specifics. I did not verify: the inventive entity of US 5,892,899 (relevant to whether it is "by another" for § 102(e)); whether Ser. No. 08/924,740 ever published or issued (if it did not, it is disclosure content by incorporation, not prior art, and is also § 103(c)-barred as commonly owned); the Aucsmith 1996 paper's exact bibliographic data; or the Davis '932 reference's content beyond the two quoting sources. Each is flagged where relied upon.
5. Date inconsistency. Tool-system date 2026‑09‑29 vs. task-stated 2026‑04‑26. Flagged, not reconciled.
6. No adjudicated § 103 holding exists for this patent. IPR2017‑01721 and IPR2020‑00213 (Unified Patents petitioner) were both terminated by settlement — per the Google Patents litigation data and the Unified Patents portal — meaning there is no institution decision and no final written decision addressing obviousness on the merits. The 2016 Federal Circuit case 16‑2700 (from the earlier Plano campaign) likewise yielded no § 103 merits ruling I could locate. Any "the PTAB already held it obvious" claim would be unsupported.

8. Bottom line

• Claims 1, 2, 19, 27, 28, 29, 31, 34, 35, 36, and 37 are highly vulnerable to § 103 over the combination of (i) admitted conditional-access/DVD art and standard public-key key transport, and (ii) tamper-resistant/obfuscated executable code with IVK-based integrity and bilateral authentication. Motivation is supplied by the patent's own statement of the problem and by the identical field and purpose of the references.
• Claims 5–8, 15–18, 21–26, 32, 33 fall to the further combination with node-authentication-then-decrypt art (US 5,633,932) and per-title key/delivery art; claim 18's billing fields are conventional commerce.
• Claims 9–14 are likely obvious but are the hardest to prove using only non-commonly-owned art, because they recite the specific IVK-generation + manifest-parser + key-compiler pipeline. This is where an obviousness attack needs the strongest published evidence (Aucsmith 1996 and the tamper-resistant-code literature) rather than the commonly-owned Intel applications.
• The single biggest practical hurdle is pre-AIA § 103(c), which bars § 103 use of the incorporated Intel applications if they are § 102(e)-only art. The attack must be reframed around (a) the specification's own admissions and (b) printed publications.
• The patent's genuine inventive kernel — deliberately transmitting an ephemeral, content-specific private key inside a tamper-resistant executable, against a background doctrine that private keys are never shared — is a real argument, but under KSR/Gurley it is more likely to preserve a narrow claim than to save claims 1 or 34.
• The expired status (2017‑12‑18) makes this largely a historical-damages question; I found no 2026 docket activity.

Sources

• Google Patents US5991399A — full text, claims, definitions, assignments, litigation, "Cited By (422)": https://patents.google.com/patent/US5991399A/en
• Unified Patents PTAB records (both settled): https://portal.unifiedpatents.com/ptab/case/IPR2020-00213 ; https://portal.unifiedpatents.com/ptab/case/IPR2017-01721
• EPO ISR citing US 5,991,399 as a "Y" reference (claims 1, 2, 4–7): https://patentimages.storage.googleapis.com/b3/d3/9f/316b94b580cfbb/WO2004019189A1.pdf
• JP search report citing US 5,991,399 as "Y": https://patentimages.storage.googleapis.com/38/88/78/3f67cc063c78b9/JP2005536797A.pdf
• US 5,633,932 (Davis et al., issued 1997‑05‑27): https://uspto.report/patent/grant/5633932 ; https://patents.searchlight.law/doc/US5633932
• Conflicting attribution listing 5224163 | June 29, 1993 | Gasser et al.: https://patents.justia.com/patent/[6898576](/patent/6898576) ; https://ia600900.us.archive.org/12/items/DTIC_ADA423264/DTIC_ADA423264.pdf ; versus US 5 224 163 A (KAUFMAN CHARLES W ET AL) in the WO 2004/019189 ISR above.