US 5974550

I'll verify current records for US 5974550 before summarizing.

Let me check the litigation/PTAB history and any 2026 appellate activity specifically for this patent.

US Patent 5,974,550 — Verification Summary

Bottom line up front: I confirmed the bibliographic and claim data for US 5,974,550 (application 08/989,615). I found no evidence of any 2026 CAFC docket activity for this patent. The only Federal Circuit reference I could find is case 16-2700 listed in Unified Patents' litigation data for the family; I could not confirm its subject matter or outcome, and I have no authoritative 2026 docket information. Treat any "2026 appeal" question as unresolved rather than negative.

Bibliographic data (all from the authoritative patent text and family records)

Assignee chain (recorded reassignments): Intel Corporation → Convera Corporation (2001-07-23) → H2 Partners LLC (2011-01-20) → Plano Encryption Technologies LLC (2015-04-03) → Honeyman Cipher Solutions LLC (2019-03-08, the current assignee listed by Google Patents). Note the entity holding the patent during the 2015–2017 assertion campaign was Plano Encryption Technologies LLC, not the current owner.

Family: WO1999031841A1, EP1042882A4/A1, JP2002509388A, KR20010032955A, AU2085699A, TW431105B (all claiming priority to 08/989,615).

Litigation / PTAB (as listed in the record): Numerous district-court cases in E.D. Tex., W.D. Tex., N.D. Tex., and D. Del., plus one Federal Circuit entry (16-2700). PTAB IPR2017-01783, filed by State Farm Mutual Automobile Insurance Company against Plano Encryption Technologies on 2017-07-14, was terminated–settled on 2019-01-09… correction: 2018-01-09 (Joint Motion to Dismiss filed 2018-01-05 under 35 U.S.C. § 317(a); termination decision 2018-01-09). A companion proceeding, IPR2017-01721, concerned related U.S. 5,991,399. Related district litigation was Plano Encryption Techs., LLC v. Alkami Tech., Inc., Nos. 2:16-cv-01032 / 2:16-cv-01072 (E.D. Tex.).

Abstract (verbatim substance)

Authenticating a remote process operating in an address space different than that of a local process includes: creating, by the local process, a tamper resistant module containing a temporary secret; sending the tamper resistant module and a challenge to the remote process; executing the tamper resistant module by the remote process and recovering the secret when the integrity of the remote process is verified by the module; encoding the challenge using the secret to produce a response; sending the response to the local process; and decoding the response by the local process. Optionally, the module includes a request for information and the response includes the answer.

Independent claims — plain-language overview

There are nine independent claims: 1 (method), 10, 14, 18, 19 (apparatus), and 20, 21, 22, 23 (machine-readable medium).

• Claim 1 (method, the core invention): A second process makes a tamper resistant module carrying a secret, sends it along with a challenge to a first process running in a different address space. The first process runs the module; only if the module verifies the first process's integrity does it recover the secret. The first process then encodes the challenge with the secret to form a response and returns it; the second process decodes it. Success means the responder is both authentic and untampered. (Dependent claims 2–9 add: integrity verification kernel; single-use secret; nonce of no persistent worth; timeout → non-authentic finding; verification by comparing a computed digital signature to a signed manifest; the challenge carrying a request for information; and encoding/decoding the answer.)

• Claim 10 (apparatus, verifying side): A machine (processing unit + storage medium) whose instructions implement the challenger's role: create the tamper resistant module with a secret, create a challenge, send both to the first process, receive the response, and decode it.

• Claim 14 (apparatus, responding side): A machine whose instructions implement the responder's role: receive the module from the second process, start executing it, recover the embedded secret when the first process's integrity is verified during module execution, receive the challenge, encode it with the secret, and send the response back.

• Claim 18 (apparatus, combined roles): An apparatus whose instructions perform both sides — create the module/secret/challenge and send them, initiate execution of that module in the address space of the first process, recover the secret upon integrity verification, receive the challenge from the second process, encode and send the response, then receive back and decode the response. Read literally, this claim recites a single instruction set that spans both endpoints; that unusual structure is worth flagging if claim scope is being analyzed.

• Claim 19 (apparatus, bilateral authentication): A local process both verifies and is verified: it (a) creates a first module with a first secret plus a first challenge, sends them to a remote process, and decodes the returned first response; and (b) receives a second module from the remote process, executes it, recovers a second secret when the local process's own integrity is verified, receives the remote's second challenge, encodes it with the second secret, and returns the second response.

• Claim 20 (machine-readable medium, challenger side): Instructions for creating the module-with-secret, creating a challenge, sending both to a remote process, receiving the response, and decoding it.

• Claim 21 (machine-readable medium, responder side): Instructions for receiving a module, executing it, recovering the secret when the local process's integrity is verified, receiving the challenge, encoding it with the secret, and returning the response.

• Claim 22 (machine-readable medium, combined roles): The medium-borne analogue of claim 18 (both challenger and responder functions in one instruction set).

• Claim 23 (machine-readable medium, bilateral): The medium-borne analogue of claim 19 (mutual authentication in both directions).

Points worth noting

1. Terminology differs between the US granted claim 1 and the PCT publication. US claim 1 recites a "tamper resistant module," whereas the corresponding WO 99/31841 claim 1 recites only a "module," with tamper-resistance moved into a dependent claim (WO claim 2). Same for WO claims 10/11, 14/15. This is a real textual difference in the documents, not a transcription error.
2. Expiry. The patent's term ran from the 1997-12-12 filing date; the record shows "Anticipated expiration 2017-12-12" and status "Expired – Lifetime." Any enforcement activity after that date would be unusual, though laches/damages questions for earlier periods are separate.
3. The cited prior art of record was sparse: US 4,521,846 (IBM), US 4,813,393 (Bull), US 4,964,163 (Motorola), and US 5,311,594 (AT&T Bell Labs). The IPR petition additionally relied on Chang references (EP 0 686 906 A2 / US 5,724,425) and U.S. RE42,762, but that proceeding settled before an institution or final written decision on the merits.

Uncertainty statement

• I have high confidence in the bibliographic fields and claim text, which come from the full authoritative patent document.
• I have low confidence / no authoritative information regarding any 2026 CAFC activity for this patent. My searches surfaced no 2026 appellate docket entry, and the single Federal Circuit reference (16-2700) in the family litigation list is unverified as to case name, outcome, and relationship to '550.
• I am not providing a legal conclusion on validity, enforceability, claim scope, or ownership. The claim summaries above are plain-language paraphrases, not constructions, and claim 18/22's combined-role structure in particular would require the prosecution history to interpret reliably.

I'll research litigation involving US patent 5974550 using the sources you specified.

Let me search for more specific litigation details.

Let me gather additional details on the Honeyman Cipher litigation and the Federal Circuit outcome.

Litigation Involving U.S. Patent No. 5,974,550

Important caveats up front: I was able to retrieve substantial data via web search (PTAB filings, RPX litigation news, the Stanford NPE Litigation Database, CourtListener, and the patent-owner's mandatory notices), but I could not fully verify every case directly on the Unified Patents litigation portal, and I exhausted my search budget before confirming all current statuses. Treat all statuses below as of the sources retrieved. Also note the patent's anticipated expiration date was 1997-12-12 + 20 years = 2017-12-12, and Google Patents shows it as "Expired – Lifetime," which is consistent with most suits having wound down.

Ownership history (relevant to who sued whom)

• Original assignee: Intel Corporation (assigned 1997-12-12; inventor Richard L. Maliszewski)
• → Convera Corporation (2001-07-23)
• → H2 Partners LLC (2011-01-20)
• → Plano Encryption Technologies LLC ("PET") (2015-04-03) — a Texas NPE, wholly owned subsidiary of Koba Holdings, LLC, run by Bradley D. Liddle
• → Honeyman Cipher Solutions LLC (2019-03-08)

The '550 patent was asserted as one of two patents (the other being U.S. 5,991,399, with U.S. 6,587,858 added later) in a single litigation campaign run by PET.

Infringement suits filed by Plano Encryption Technologies (as plaintiff)

The '550 patent was specifically identified in the infringement complaint against State Farm (Count Two, asserting at least claim 20), and in the complaints against mobile-banking defendants (e.g., claims 14–17).

Declaratory judgment actions (banks/software vendors as plaintiffs, PET as defendant)

PTAB and appellate proceedings (specific to the '550 patent)

• IPR2017-01783 — State Farm Mutual Automobile Insurance Company v. Plano Encryption Technologies LLC (et al.), PTAB. Petition filed 2017-07-14; challenged U.S. 5,974,550. Status: Terminated – Settled (termination decision 2018-01-09; petitioner's motion to dismiss and settlement agreement filed 2018-01-05). Respondent patent number listed as 5,974,550; application 08/989,615; Tech Center 2700. Petitioner counsel: R. William Beard, Jr.; Patent Owner counsel: Bradley Liddle.
• IPR2017-01721 — State Farm Mutual Automobile Insurance Company v. Plano Encryption Technologies, LLC (PTAB, filed 2017) — a companion State Farm IPR in the same campaign (identified in the same PTAB filings).
• Jack Henry & Associates, Inc. v. Plano Encryption Technologies, No. 16-2700 (Fed. Cir.) — appeal of the N.D. Tex. declaratory judgment action. Decided as 910 F.3d 1199 (Fed. Cir. 2018) (venue/personal-jurisdiction issues). The CourtListener record confirms the '550 patent (claims 14–17 cited in the underlying claim charts) was at issue.

Later owner (Honeyman Cipher Solutions LLC)

After acquiring the portfolio in 2019, Honeyman Cipher launched its own campaign (e.g., Honeyman Cipher Solutions LLC v. Groupon, 1:19-cv-03754, N.D. Ill.; v. Snap, 1:19-cv-01547, D. Del.; v. LogMeIn, 1:19-cv-01545, D. Del.; and 2020 suits against Adidas, ASICS, Evernote, PayPal, Asana, Slack, IKEA, etc.). However, the reporting I retrieved indicates the Honeyman Cipher complaints asserted the '399 patent, not the '550 patent, and the Stanford NPE database entry for 5,974,550 does not list Honeyman Cipher cases. I therefore could not confirm any Honeyman Cipher assertion of the '550 patent specifically.

Bottom line

Yes — extensive litigation involves U.S. 5,974,550: roughly 14 infringement suits and 5 declaratory judgment actions, almost all brought by (or against) Plano Encryption Technologies LLC in E.D. Tex., W.D. Tex., N.D. Tex., and D. Del. between 2015 and 2017, targeting banks, online/mobile-banking software providers, and consumer mobile-app operators. There was also one settled inter partes review (IPR2017-01783) and one Federal Circuit appeal (16-2700 → 910 F.3d 1199). The patent expired (anticipated) on 2017-12-12. The later Honeyman Cipher campaign appears to have asserted the sibling '399 patent rather than the '550.

Key sources: IPR2017-01783 (GreyB/IPVerse); PET Patent Owner Mandatory Notice, IPR2017-01783 (USPTO PTAB); Stanford NPE Litigation Database – 5974550; RPX Empower litigation news; Broadway National Bank v. PET, CourtListener; Jack Henry v. PET, 910 F.3d 1199, CourtListener; Unified Patents – Fidelity v. Koba/PET (1:15-cv-00777).

Note: I am a patent analyst, not your attorney, and this is not legal advice; statuses should be re-verified against PACER/Unified Patents before being relied upon.

I'll research the PTAB records for this patent before writing up.

Proceedings overview

One AIA trial proceeding on file for US 5,974,550 — IPR2017-01783 — and it was terminated on settlement before the Board ever decided whether to institute. Breakdown: 1 total / 0 active / 0 claims invalidated / 0 claims sustained / 1 settled / 0 institution denials. The bottom line for a defendant is therefore unusually flat in both directions: the patent is not "hardened" by any PTAB win, but it is also not narrowed — all 23 claims stand exactly as issued, untested, and the patent expired on 2017-12-12. Nothing the PTAB did helps you kill the claims, and nothing it did hurt them either; the expiry date, not the PTAB record, is the dispositive fact.

⚠️ Contradiction flagged. The "PTAB proceedings on file" block supplied with this prompt states the USPTO Open Data Portal returns no AIA trial proceedings for this patent. That is wrong for this patent. The proceeding IPR2017-01783 is corroborated by (a) the full patent record itself (Legal Events entry, 2017-08-29: "Aia trial proceeding filed before the patent and appeal board: inter partes review — TRIAL NO: IPR2017-01783, Opponent name: STATE FARM MUTUAL AUTOMOBILE INSURANCE COMPANY"), (b) the Board's own 2018-01-09 termination judgment fetched from docketalarm, and (c) contemporaneous PTO litigation reporting. I treat the structured block as an ODP indexing gap for a terminated pre-institution proceeding, not as evidence of no activity. Do not rely on the "no PTAB activity" default for this patent.

IPR2017-01783 — State Farm Mutual Automobile Insurance Company v. Plano Encryption Technologies, LLC

• Type: Inter Partes Review (35 U.S.C. §§ 311–319; pre-AIA §§ 102/103 — patent priority 1997-12-12)
• Filed: 2017-07-14 (filing date accorded by Notice mailed 2017-08-07; petition assigned Paper 1)
• Status: Terminated – Settled (verbatim per proceeding records); plain English: never instituted, never decided on the merits. The Board terminated at the preliminary proceeding stage — "the Board has not decided whether to institute a review."
• Judge panel: Joni Y. Chang (writing), Annette R. Reimers, Michelle N. Wormmeester — the panel that entered the termination judgment. No panel ever reached the merits.
• Petition grounds (as recited in Patent Owner's Preliminary Response, Paper 6, 2017-11-07): cancellation of claims 10, 11, 12, and 20 under 35 U.S.C. § 103, on two grounds:
  • Ground 1 — obvious over Aucsmith (Tamper Resistant Software) in view of Chang (EP 0 686 906 A2, Ex. 1010; and U.S. 5,724,425, Ex. 1011);
  • Ground 2 — same combination further in view of the knowledge of a person of ordinary skill in the art.
  • Support exhibits included Ex. 1003 (Expert Declaration of David I. August, Ph.D.), Ex. 1012 (Declaration of Scott Bennett, Ph.D.) (prior-art/publication authentication), Ex. 1017 (U.S. RE42,762, "Shin"), plus the '550 and '399 file wrappers and the E.D. Tex. claim-construction record. Claim-construction dispute focused on "tamper resistant module" (and the absence of any construction for "integrity verification kernel").
• Institution decision: None — no institution decision was ever issued. The proceeding terminated first. (For comparison, the companion petition IPR2017-01721 against related U.S. 5,991,399, filed 2017-07-13 by the same petitioner, was likewise terminated before institution on the same day.)
• Final Written Decision: None issued. No claim of the '550 patent has ever been canceled, confirmed, or otherwise adjudicated by the PTAB. Do not credit any statement that a claim of this patent was "invalidated in IPR" — the Board never reached the merits.
• Settlement / termination: Joint Motion to Terminate filed 2018-01-05 (Paper 7) under 35 U.S.C. § 317(a) and 37 C.F.R. § 42.73, supported by a true copy of the written Settlement Agreement (Ex. 1019). The parties represented the settlement "resolves all disputes between the parties as to the involved patents," and the related district court actions were dismissed (Ex. 1020, Order of Dismissal). The Board granted termination on 2018-01-09 and terminated the proceeding as to all parties. Terms are confidential — the settlement agreement was filed as an exhibit, but nothing in the public record discloses consideration, licensing terms, or covenants. A Notice of Refund mailed 2018-01-12 returned $14,000 in post-institution fees to State Farm, consistent with the case never reaching trial.
• Appeal: None. There was no FWD, therefore no appealable Board decision. (Note: the only Federal Circuit entry in this family, Jack Henry & Assocs., Inc. v. Plano Encryption Techs., LLC, No. 2016-2700, 910 F.3d 1199 (Fed. Cir. 2018-12-07) — precedential, Newman/Wallach/Stoll — is not an appeal of any PTAB decision. It is a personal-jurisdiction/venue appeal reversing dismissal of a declaratory-judgment action in N.D. Tex.; the asserted letters cited "at least claims 14–17 of U.S. Patent No. 5,974,550.")
• Defensive value: Low, but not zero. There is no claim-level kill shot to inherit and no § 315(e) estoppel to worry about. What you do inherit is the petitioner's work product: a fully briefed Aucsmith + Chang § 103 theory against claims 10, 11, 12, and 20, developed with an expert declaration, that was never tested by the Board — the Board never accepted or rejected it. See the strategic section below for why that ground is still technically open to a new petitioner even though the patent is expired.

Strategic summary

Claim status: no claim is canceled; no claim is sustained; all 23 claims are UNTESTED at the PTAB. The complete claim set (1–23) was issued in 1999 and remains unamended and unadjudicated by the Board. The proceeding that existed — IPR2017-01783 — challenged only claims 10, 11, 12, and 20 and died in the preliminary stage. For anyone facing a demand letter on this patent, that means: (i) you cannot say "claim 1 is dead" — claim 1 was never even challenged; (ii) you cannot say "the patent survived PTAB scrutiny" — nothing was scrutinized; and (iii) the only genuinely decisive fact is that the patent expired 2017-12-12, per the record's "Anticipated expiration 2017-12-12" and "Expired – Lifetime" status. Enforcement of an expired patent is limited to past damages within the 35 U.S.C. § 286 six-year lookback, and even that presupposes a live suit (no current assignee litigation appears in the record; the current owner of record is Honeyman Cipher Solutions LLC as of 2019-03-08, having acquired from Plano Encryption Technologies LLC).

Estoppel landscape: essentially a blank slate, which cuts both ways. Section 315(e)(1)/(2) estoppel attaches to a petitioner only after an IPR that has been instituted and carried to a final written decision. Here there was no institution and no FWD, so State Farm is not estopped on the Aucsmith + Chang combination or anything else, and neither is anyone else. That also means the Aucsmith + Chang ground remains available to a fresh petitioner — a new challenge would not be barred by IPR2017-01783 as a matter of § 315(e), though a real-world petition would have to contend with the Board's § 325(d) discretion (same or substantially the same art/arguments previously presented) and with the § 315(b) one-year bar running from service of any complaint. Note the practical ceiling: an IPR against an expired patent is procedurally possible (the Board can institute after expiration; Patent Owner itself cited Arris Group Inc. v. C-Cation Technologies LLC, IPR2014-00746, on the standard to apply), but there is no injunctive upside and no amendment path for the patent owner — it is a pure validity-cleansing exercise with diminishing economic value now that the term has run.

Pattern signals: this was a defendant-by-defendant, one-off challenge, not a coordinated attack. The sole petitioner was State Farm, a litigation defendant facing Plano Encryption in Plano Encryption Techs., LLC v. State Farm Mut. Auto. Ins. Co., No. 2:16-cv-01072 (E.D. Tex.), and it filed a matched pair of petitions on the same day (IPR2017-01783 on '550; IPR2017-01721 on '399) and then bought its way out of both by 2018-01-05, roughly six months after filing and before any institution decision. No defensive aggregator attacked this patent — Unified Patents did file an IPR against Plano Encryption (IPR2016-01697), but on a different patent, U.S. 9,044,226, so there is no Unified Patents chain here. The patent owner (a single-officer Texas NPE whose CEO and corporate counsel were the same person, per RPX's 2016 campaign reporting) never had occasion to defend a PTAB appeal, because it never had an adverse FWD to appeal. The result is an unusually thin PTAB record for a patent that was asserted against roughly two dozen defendants across E.D./W.D./N.D. Tex. and D. Del. in the 2015–2016 campaign.

Recommended next steps

• If you have a demand letter or suit asserting US 5,974,550, the threshold message is expiry, not PTAB: the term ran from the 1997-12-12 filing date and the record shows anticipated expiration 2017-12-12. Confirm the fee-status history (the Legal Events show maintenance fees paid at 4, 8, and 12 years, plus a 2011-11-30 late-payment surcharge) so that no lapse argument is missed, but assume the patent is expired.
• Do not overstate the PTAB record in either direction. The correct citation is the Board's termination judgment of 2018-01-09 in IPR2017-01783, which expressly states that "the Board has not decided whether to institute a review" and terminates the preliminary proceeding. Cite it for the proposition that no claim was adjudicated. Full text: https://www.docketalarm.com/cases/PTAB/IPR2017-01783/Inter_Partes_Review_of_U.S._Pat._5974550/docs/01-09-2018-Board/Termination_Decision_Document-8-Termination_Decision_Document.pdf (proceeding record: https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1494394](/patent/1494394); mirrored at https://ipverse.greyb.com/ptab-web/cases/case-details/IPR2017-01783).
• If you are considering an affirmative PTAB challenge: there are no active proceedings, so there are no trial-stage milestones to track. Any new petition would be on an expired patent, subject to the § 315(b) one-year bar, and would face § 325(d) discretion in light of the Aucsmith/Chang art already presented in IPR2017-01783. The petitioner's own expert declarations (August, Ex. 1003; Bennett, Ex. 1012) are public and reusable as a starting draft — but note the Board never blessed that theory, so treat it as unvalidated work product.
• For the invalidity/claim-scope fight, look to the district court record, not the PTAB. The relevant construction history is Plano Encryption Techs., LLC v. American Bank of Texas, LLC, No. 2:15-cv-01273-JRG (E.D. Tex.), 2016 WL 3959808 (2016-07-22) — the construction that Patent Owner argued judicially estopped State Farm from taking contrary positions in the IPR. That order, not anything the Board did, is where the operative claim-scope arguments about "tamper resistant module" and "integrity verification kernel" live.
• Standing/jurisdiction note for any renewed dispute: if a successor owner reactivates this family, Jack Henry & Assocs. v. Plano Encryption Techs., 910 F.3d 1199 (Fed. Cir. 2018-12-07), is directly on point and adverse to the patent owner on the proposition that infringement/licensing letters cannot, as a categorical rule, avoid personal jurisdiction — it also confirms that an indemnifying supplier has standing to participate. Opinion: https://www.courtlistener.com/opinion/[4571118](/patent/4571118)/jack-henry-associates-inc-v-plano-encryption-technologies/.

Confidence note: High confidence on the existence, parties, dates, challenged claims, grounds, panel, and disposition of IPR2017-01783, and on the absence of any FWD or appeal — all corroborated by Board documents. I found no evidence of any other AIA proceeding (IPR, PGR, CBM) or ex parte reexamination on US 5,974,550; that negative is stated as my search result, not as a certified USPTO clearance. I render no opinion on validity, enforceability, or infringement.

I'll verify what I can from the assignment record and search for correspondent-of-record details before assembling the chain.

Let me dig into the correspondents and NPE principals for each assignee.

Assignment & Ownership-Chain Analysis — US 5,974,550

Access caveat up front: I do not have a live, authenticated connection to USPTO Assignment Center and my searches cannot render its results table. The reel/frame numbers below are taken from the Google Patents "Legal Events" record for US 5,974,550 (which reproduces the assignment-reel citations), and I corroborated the substance of the 2015 and 2019 links against primary documents filed in IPR2017-01783. The correspondent-of-record field was not retrievable for any link — I flag that explicitly rather than guess. Verify at https://assignmentcenter.uspto.gov/ (search by patent number 5974550).

Inventors

Pattern note: No unusual inventor patterns. The assignment was executed 1997-12-10, two days before the 1997-12-12 filing date — an ordinary at-filing employment assignment, not a post-filing cleanup. There is no co-inventor attrition, no license-back, and no inventor-retained interest. The '550 is a single-inventor patent; the sibling Intel tamper-resistance filings referenced in the specification (Ser. Nos. 08/662,679 and 08/924,740) are separate applications, and I have not verified their inventorship. No red flags on the inventor row.

Original assignee

Intel Corporation (Santa Clara, California) — named on the issued patent and as original assignee per the 1997-12-10 assignment (reel 008917/0546).

• Primary line of business: semiconductor design and manufacturing — microprocessors (the specification expressly names PENTIUM® and PENTIUM® II, its own products, as the target platforms).
• Did Intel ship a product embodying the claims? No — and this matters. The claimed subject matter is a software challenge-response/tamper-resistant-module authentication protocol. Intel's business in 1997 was silicon, not packaged authentication software. The specification describes the invention generically (two processes, "may or may not exist on the same processor or computer system") and cites Intel's own 1997-era tamper-resistant-compiler work, but I found no evidence of a commercial Intel product practicing the claims. Reliability caveat: I am asserting a negative from absence of evidence, which is weak; Intel's internal use of tamper-resistant object code (the "Tamper Resistant Methods and Apparatus" work) could conceivably practice it.
• Current status: Operating (Intel Corporation). Also relevant to the chain: Convera Corporation (the 2001 assignee) was an enterprise-search software company formed from the combination of Excalibur Technologies and Intel's Interactive Media Services division, with Intel holding an equity stake — i.e., the 2001 transfer was an affiliate/spinout transfer, not an arm's-length sale to a stranger. Convera's search business was subsequently sold to FAST Search & Transfer (2007). I am moderate-confidence on the Convera corporate history details and high-confidence that Convera was an Intel-affiliated operating software company, not an NPE.
• Bankruptcy: No evidence of any Chapter 7/11 proceeding involving Intel, Convera, H2 Partners, or Plano Encryption Technologies in connection with this patent. Not verified either way for Convera's later corporate wind-down.

Assignment timeline

Five recorded assignments. No security agreements, no liens, no releases, no change-of-name-only records, no merger records appear in the chain.

• 1997-12-10 (executed) / recorded 1997-12-12 — Reel 008917/0546

  • Conveyance: Assignment (Assignment of Assignors' Interest)
  • Assignor: Richard L. Maliszewski
  • Assignee: Intel Corporation (California)
  • Correspondent: Not retrieved.
  • Context: Standard inventor-to-employer assignment executed two days pre-filing. Not a red flag.

• 2001-07-17 (executed) / recorded 2001-07-23 — Reel 012002/0742

  • Conveyance: Assignment (Assignment of Assignors' Interest)
  • Assignor: Intel Corporation
  • Assignee: Convera Corporation (Virginia)
  • Correspondent: Not retrieved.
  • Context: Affiliate/spinout transfer to an Intel-linked operating software company (Convera arose from Excalibur + Intel's Interactive Media Services). Not an arm's-length monetization sale.

• 2011-01-11 (executed) / recorded 2011-01-20 — Reel 025667/0929

  • Conveyance: Assignment (Assignment of Assignors' Interest)
  • Assignor: Convera Corporation
  • Assignee: H2 Partners LLC (Connecticut — New Canaan, CT per the 2015 assignment recital)
  • Correspondent: Not retrieved.
  • Context: Disposition of legacy Intel-origin assets out of a software company whose search business had already been sold (2007). This is the first link to a non-operating holder, and it is the least documented link in the chain — I could not determine whether this was a portfolio-wide divestiture, a distressed sale, or a broker-intermediated transaction. Flagging as an evidence gap, not a finding.

• 2015-03-18 (executed) / recorded 2015-04-03 — Reel 035327/0227

  • Conveyance: Assignment (Assignment of Assignors' Interest)
  • Assignor: H2 Partners LLC (New Canaan, CT)
  • Assignee: Plano Encryption Technologies, LLC (primary place of business: Plano, TX) — a Texas LLC formed March 2015, wholly owned subsidiary of Koba Holdings, LLC (Delaware), sole officer/employee Bradley D. Liddle
  • Correspondent: Not retrieved from the recorded field. However — the assignment instrument produced as an exhibit in IPR2017-01783 is associated with Bradley D. Liddle, Reg. No. 71,106, 903 E. 18th St., Suite 224, Plano, TX 75074 — the identical address PET used for service and the identical individual named as PET's CEO, corporate counsel, and PTAB lead counsel in the same IPR. Flag: Liddle recurs as the human being behind the PET-side paperwork on this link; he is also the named contact on PET's cease-and-desist ("Liddle Letter") activity. I cannot confirm he is the recorded correspondent on reel 035327/0227, only that he signs and speaks for the assignee.
  • Context: Transfer-to-asserter. The instrument conveys a four-patent package plus three applications — U.S. 5,991,399; 5,974,550; 6,041,122; 6,529,603 and apps 09/441,409; 09/603,079; 09/709,901 — together with all causes of action and enforcement rights, including past damages and injunctive relief. Fired ~3.5 months before PET's first infringement complaint. This is the marquee link.

• 2019-03-08 (executed) / recorded 2019-03-08 — Reel 048540/0020

  • Conveyance: Assignment (Assignment of Assignors' Interest)
  • Assignor: Plano Encryption Technologies LLC
  • Assignee: Honeyman Cipher Solutions LLC (Texas)
  • Correspondent: Not retrieved.
  • Context: Transfer-to-asserter (second generation). Executed and recorded the same day, ~9 months after the '550's anticipated expiration (2017-12-12) and 81 days after PET's IPR2017-01783 was terminated by settlement. The surrounding Honeyman Cipher campaign, as reported by RPX/Stanford, appears to have been driven by the sibling '399 patent rather than the '550 — which is consistent with the '550 being expired and largely spent by 2019. Caveat: I could not confirm whether the '550 was even included in this instrument; the reel citation comes from the Google Patents legal-events record, and I did not retrieve the underlying instrument.

Timeline diagram

timeline
    title Ownership of US 5974550
    1997 : Inventor assigns to Intel Corp
         : Application filed 12 December
    1999 : Patent issues 26 October
    2001 : Intel assigns to Convera Corporation
    2011 : Convera assigns to H2 Partners LLC
    2015 : H2 assigns to Plano Encryption Technologies
         : First infringement suits filed June
    2017 : Patent expires 12 December
    2019 : Plano assigns to Honeyman Cipher Solutions

NPE / troll-pattern signals

1. Shell-entity transfer — PRESENT

Reel 035327/0227 (executed 2015-03-18) moves the patent from H2 Partners LLC to Plano Encryption Technologies, LLC, a Texas LLC formed in March 2015 — i.e., the assignee was created in the same month it took title. Concrete corroboration, not naming inference:

• PET is a wholly owned subsidiary of Koba Holdings, LLC (Delaware), stated in PET's own mandatory notice in IPR2017-01783 and repeated in the State Farm settlement agreement.
• Bradley D. Liddle is PET's only employee and officer (RPX Empower).
• PET ships no products; its stated activity is licensing and litigation. Its "address" (903 E. 18th St., Suite 224, Plano, TX) is Liddle's own business address.
• A D. Delaware opinion in Fidelity National Information Services v. Plano Encryption Technologies, 1:15-cv-00777, addressed and rejected an alter-ego theory aimed at reaching Koba — meaning the two-entity structure was litigated and at minimum resisted veil-piercing.
• Reel 048540/0020 (2019) repeats the pattern: transfer to Honeyman Cipher Solutions LLC, another single-purpose Texas LLC.

2. Known asserter in the chain — PRESENT (qualified)

Neither Plano Encryption Technologies LLC nor Honeyman Cipher Solutions LLC appears on the enumerated canonical list (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg). The finding rests on the "any entity surfaced by Unified Patents or RPX as a high-frequency plaintiff" prong:

• RPX Empower documents PET's "sole litigation campaign" — ~14 infringement suits and ~5 declaratory judgment actions across E.D. Tex., W.D. Tex., N.D. Tex., and D. Del. between June 2015 and late 2016, asserting the '550 and '399 consistently from the outset.
• Unified Patents' litigation portal lists this patent in numerous district-court dockets plus IPR2017-01783.
• PET's CEO is described by RPX as having "connections to other NPEs."

3. Repeat correspondent across the chain — PRESENT (qualified) / partially unclear

I could not retrieve the recorded correspondent field for any of the five reel/frame entries, so I cannot state the classic version of this signal ("same attorney recorded on reels X, Y, Z"). What I can document from primary filings:

• Bradley D. Liddle, Reg. No. 71,106, appears as the assignee-side signatory/contact on the H2→PET instrument surfaced in IPR2017-01783, as PET's lead counsel in that IPR, as PET's service contact, and as the named address on PET's pre-suit demand letters (Broadway National Bank v. PET, W.D. Tex. 1:15-cv-01056).
• He is thus a single repeat player controlling both the asset-holding entity and the assertion/recording function for the PET links. That is exactly the function the signal is designed to catch — one lawyer behind LLCs designed to look unrelated — but note he appears on one recorded link, not several, so I stop short of calling it a cross-chain recurrence.
• No correspondent identified at all for the 1997, 2001, 2011, and 2019 links.

4. Cascading transfers — NOT PRESENT

The links are spaced 1997 → 2001 → 2011 → 2015 → 2019. Even the densest cluster (H2 → PET, 2011-01-11 to 2015-03-18) is ~50 months, well outside the <24-month window. But note the adjacent pattern: the same Intel-origin package (four patents + three applications) was re-sold as a bundle in 2015 after sitting with H2 Partners for four years — recurring monetization of one asset bundle across holders, which the <24-month test does not capture.

5. Pre-litigation transfer — PRESENT (strongest signal in the chain)

• Assignment executed 2015-03-18, recorded 2015-04-03 (reel 035327/0227).
• PET's first infringement complaint filed 2015-06-29 — Plano Encryption Technologies, LLC v. Citizens National Bank, E.D. Tex. 2:15-cv-01168-JRG.
• Interval: ~3.5 months from execution to first suit, and only ~12 weeks from recording. The assignee LLC was formed in the same month it took title. The standing record was cleaned up immediately before the campaign began.

6. Bankruptcy fire-sale — NOT PRESENT (no evidence); one gap

No Chapter 7/11 proceeding is disclosed anywhere in the assignment record, Google Patents legal events, or the IPR documents for Intel, Convera, H2 Partners, PET, or Honeyman Cipher. The 2011 Convera → H2 Partners transfer occurred roughly four years after Convera sold its RetrievalWare search business to FAST (2007), which is suggestive of a wind-down divestiture but is not a court-supervised sale and I found no filing to confirm the mechanism. Marked not present, with the 2011 link flagged as an unexplained disposition.

7. Privateering — NOT PRESENT

Intel's 2001 transfer went to Convera, an Intel-affiliated operating company — a corporate transaction, not a proxy. I found no evidence that Intel funded, controlled, directed, or shared in PET's 2015–2016 campaign, and no SEC disclosure tying Intel to it. The 2015 transfer was to an independent asserter, which is monetization, not privateering.

8. Defensive aggregator — NOT PRESENT

The chain terminates at Honeyman Cipher Solutions LLC (reel 048540/0020, 2019-03-08). No RPX, AST, LOT Network, Unified Patents, or OIN link anywhere in the record. The patent was not neutralized — it was sold onward to a second-generation asserter.

Verdict

NPE — high confidence

Two independently sufficient strong signals, each grounded in a specific reel/frame and date: (1) shell-entity transfer — reel 035327/0227 (executed 2015-03-18) moved the patent from H2 Partners LLC to Plano Encryption Technologies, LLC, a Texas LLC formed in the same month, wholly owned by Koba Holdings LLC, with Bradley D. Liddle as its sole officer and employee, zero products, and a corporate address identical to its lawyer's; and (2) pre-litigation transfer — the assignment was executed ~3.5 months before PET's first infringement complaint (2015-06-29, E.D. Tex. 2:15-cv-01168), and the instrument conveyed a four-patent bundle plus all past-infringement causes of action, i.e., a transaction purpose-built for assertion. Reinforcing it, reel 048540/0020 (2019-03-08) shows a second transfer-to-asserter to Honeyman Cipher Solutions LLC, and RPX/Unified record PET's campaign at roughly 14 infringement suits and 5 DJ actions — squarely within the "high-frequency plaintiff surfaced by RPX/Unified" prong.

Two honest limits on this verdict: (a) the recorded correspondent field could not be retrieved for any link, so the repeat-correspondent signal is evidenced only indirectly (Liddle's role in the IPR and the assignment instrument) rather than from the assignment records themselves; and (b) this patent was held by an operating company (Intel) or its affiliate (Convera) for its first 13 years — the NPE character attaches only from the 2011 H2 Partners link onward, and the 2011 mechanism is the least documented step in the chain.

Verification link: https://assignmentcenter.uspto.gov/ (search patent number 5974550); mirrored at https://assignment.uspto.gov/patent/index.html

Cross-reference notes against the previously generated sections

• Reconciled, no contradiction: the prior "Litigation summary" listed PET's acquisition as 2015-04-03, which is the recording date; the execution date is 2015-03-18 (Google Patents legal events, "Effective date: 20150318"). Both are correct; the distinction is material for the pre-litigation-transfer calculation above.
• Builds on, not repeats: the earlier sections already established the Intel → Convera → H2 → PET → Honeyman chain and PET's Koba/Liddle structure. This section adds the reel/frame citations, the executed-vs-recorded dates, the bundle-transfer scope with past-damages rights, and the correspondent evidence gap.
• New adverse fact for the prior sections' litigation narrative: the State Farm settlement agreement text (IPR2017-01783 record) recites that the E.D. Tex. court granted summary judgment of non-infringement as to the '550 Patent before the parties settled. The earlier litigation summary reported IPR2017-01783 as terminated-settled but did not surface the '550 non-infringement SJ — worth adding to that section.
• One inconsistency in my own prior output, now corrected: the earlier "Litigation summary" gave the Jack Henry DJ action as 3:15-cv-03745-N; the Broadway/PET motion papers quote 3:15-cv-3645-N. Reel/frame data does not resolve this; PACER should.

This is patent-prosecution analysis, not legal advice; assignment statuses should be re-verified against Assignment Center and PACER before reliance.

I'll research the prior-art citations for US 5,974,550 using the USPTO and other authoritative sources.

Let me research the remaining cited references and the IPR petition's prior-art grounds.

Let me examine the IPR petition's prior-art grounds in detail.

I've hit my search-step limit, so here is the analysis based on the authoritative patent document (which lists the citations of record) and the sources retrieved. I'll flag where I could not verify something.

Prior Art for US 5,974,550 — Analyst Report

0. Verification of the target patent

I confirmed US 5,974,550 on the USPTO/Google Patents record: Method for strongly authenticating another process in a different address space, app. 08/989,615, filed 1997-12-12, issued 1999-10-26, inventor Richard L. Maliszewski, original assignee Intel Corp., 23 claims, status Expired – Lifetime. This is the correct number; I did not return near-miss numbers.

Critical framing before the reference-by-reference analysis:

The '550 patent's own "References Cited" list contains only four U.S. patent references, all cited by the examiner. That is an unusually thin art-of-record for a 1997 software-security case, and it materially affects how the §102 question should be answered (see §4). There is also a separate list, "Family Cites Families (6)," which are references cited in the international/foreign search reports for the family — not the U.S. examiner. I treat the two groups separately, because only the first group constitutes the art of record for the U.S. patent.

All four examiner-cited references were filed/published more than one year before the 1997-12-12 filing, so each is presumptively available as 35 U.S.C. § 102(b) prior art (patents/printed publications dated more than one year before the U.S. filing date). That is the correct statutory basis; none should be analyzed under §102(a)/(e) given these dates.

1. The four examiner-cited references (art of record)

Sources: Google Patents US5974550A, US4811393A, US5311594A, uspto.report US4521846, Unified Patents US-4811393-A.

Reference A — US 4,521,846 (IBM)

• Full citation / dates: US 4,521,846, Scalzi et al., app. 06/236,387, priority 1981-02-20, granted 1985-06-04, IBM. (EP counterpart EP 0 058 844.)
• Brief description: A System/370 cross-memory architecture. A general-purpose-register (GR) mask associates predetermined address spaces with specific GRs used as intra-address-space base registers; an address-space mask register selects among multiple simultaneously available address spaces designated by STO registers, with storage-protect keys controlling authorized access. It lets an executing program access data in plural address spaces in non-privileged state (including storage-to-storage instructions).
• §102 analysis — which claims could it anticipate? None. The reference is cited for the environment only: it is background showing that "a process operating in an address space different than that of a second process" was known technology decades earlier. It discloses no tamper resistant module, secret, challenge, response, or integrity verification. It cannot anticipate claim 1, 10, 14, 18, 19, or 20–23 because it lacks every substantive limitation after the preamble. At most it is relevant to the preamble of claim 1/10/14 (two processes in different address spaces) — a preamble alone does not anticipate.

Reference B — US 4,811,393 (Bull)

• Full citation / dates: US 4,811,393, Hazard, app. 07/171,877, priority FR 1986-07-16, granted 1989-03-06; assignee Bull, S.A.
• Brief description: Key diversification and authentication for memory cards. An initialization system derives a diversified key Sd from a basic key Sb via a biunique combinatorial transformation; at use, an exploitation system computes a "certificate" R1 = f3(K1, Ex) from the basic key while the card computes R2 = f4(K2, Ex) from its diversified key. The transaction is authorized only if the two certificates match. Claim 1 recites an external datum (Ex) serving as a challenge to which the card must produce a matching response.
• §102 analysis: This is the most conceptually relevant of the four to the cryptographic core of the '550 — a secret key plus an external datum (Ex) producing a response that the verifier checks. It is genuinely relevant to:
  • claim 1 — only insofar as "encoding the challenge using the secret to produce a response; … decoding the response" (the keyed challenge-response concept), and
  • claim 6 — the "compare a computed value to a predetermined value" idea (here, certificate equality), though not a signed manifest of a code image.
  • However, it does not anticipate any claim. It lacks the tamper resistant module, the "recover the secret when the integrity of the first process is verified by the module" limitation, and the different-address-space process limitation. It is a card/terminal cryptographic protocol, not an agent-based integrity-verification protocol. Best characterized as §102(b) art that is relevant to individual limitations, not anticipatory.

Reference C — US 4,964,163 (Motorola)

• Full citation / dates: US 4,964,163, app. filed 1988-04-04, granted 1990-10-16, Motorola, Inc. (same-title continuation family includes US 5,274,368.)
• Brief description: Access control in a radio/telephone communication system. A central unit transmits a "Puzzle" to a subscriber unit; the subscriber solves the puzzle (operating on one or more digital codes with operand codes) to derive an "Answer" code and transmits it back; the central unit returns an "Acknowledge" if the answer is correct, or denies access. The subscriber routine includes a time-out/retry mechanism (decision 110 → retransmit decision 112) if no response arrives within a time interval.
• §102 analysis: This is the closest of the four to the challenge-response and timeout limitations, and is the reference most plausibly argued:
  • claim 1 — the "challenge → response → verify" structure (Puzzle → Answer → Acknowledge), but no secret embedded in a module that releases only on integrity verification, and no address-space element;
  • claim 5 — determining that the first process is not authentic when the response is not received within a predetermined period of time — the reference's retry-after-timeout logic is materially related to this limitation.
  • Does it anticipate? No. It lacks the tamper-resistant module, the integrity-verification-released secret, and the different-address-space process; and its "puzzle/answer" is an arithmetically self-contained computation, not a keyed encoding using a verifier-known secret.

Reference D — US 5,311,594 (AT&T Bell Labs / Penzias)

• Full citation / dates: US 5,311,594, A. A. Penzias, filed 1993-03-26, granted 1994-05-10, AT&T Bell Laboratories.
• Brief description: Fraud protection for card transactions via content-based challenge-response. A security gateway prestores several pieces of personal information per account; for each transaction it randomly selects one piece (and random digit positions within it) and asks the caller to supply it. Access is granted only if the supplied digits match. The authentication information is randomly determined for each transaction, defeating replay of the previously-used response.
• §102 analysis: Relevant to the "fresh challenge per transaction" idea underlying claim 1 and to claim 4 (the secret is a nonce of no persistent worth) by analogy only — the reference's randomness is in the challenge selection, not a per-transaction secret. It does not anticipate any claim: there is no module, no secret released on integrity verification, no cryptographic encoding, and the "response" is human knowledge, not a cryptographic function of a secret.

2. The "Family Cites Families" references (international/foreign search-report citations)

These six are cited in the family's foreign search reports, not by the U.S. examiner. Two of them post-date the '550 priority date and therefore are not §102 prior art against the U.S. patent (they could only matter abroad, against the later-filed PCT/foreign cases).

3. The references actually asserted in the IPR (the only true invalidity theories)

The only proceeding that asserted specific anticipation/obviousness grounds against the '550 was IPR2017-01783, State Farm Mutual Automobile Insurance Co. v. Plano Encryption Technologies LLC, petition filed 2017-07-14. Per the petition documents I retrieved, the grounds were obviousness combinations, not single-reference anticipation — e.g., "CLAIMS 10, 11, 12, AND 20 OF THE '550 PATENT ARE OBVIOUS OVER AUCSMITH IN VIEW OF CHANG," and a parallel chart for claims 14–17. References relied on:

• "Aucsmith" (Ex. 1009) — a tamper-resistant-software disclosure describing IVKs (including an "eIVK") that "authenticate themselves, the other IVKs, and other programs… through the Integrity Verification Protocol Program," with "challenge/response authentication between the IVK elements." This is the closest prior art to the '550's tamper-resistant-module-with-IVK concept. I could not verify Aucsmith's exact document type/number within my search budget — flagging as unverified.
• "Chang" — identified in the earlier-generated section as EP 0 686 906 A2 / US 5,724,425 (challenge-response authentication of access rights). Moderate confidence; corroborated by the DE 69704684 T2 family member above.
• U.S. RE 42,762 — also relied on in the petition.

Outcome: the IPR was terminated–settled (2018-01-05/09) before institution. There is therefore no institution decision and no final written decision on the merits, meaning no PTAB finding that any claim of '550 is anticipated or obvious. The petition's own characterizations of Aucsmith and Chang are advocacy, not adjudicated findings. (PTAB petition document, USPTO PTACTS)

4. Direct answer: which claim(s) does each citation "potentially anticipate" under §102?

Being precise, as the analyst should be:

• No cited reference — none of the four examiner citations and none of the six family citations — anticipates any claim of US 5,974,550 in full under §102. Anticipation requires a single reference disclosing every limitation arranged as in the claim. Every asserted claim (independent claims 1, 10, 14, 18, 19, and 20–23) requires, at minimum: (i) a tamper resistant module containing a secret, (ii) the module being executed by the other process, and (iii) recovery of the secret contingent on the executing module verifying that other process's integrity. None of the four cited references discloses element (i) or (iii) in any form.
• The citations are therefore background/limitation-level art, mapping as follows:
  • US 4,521,846 → preamble of claims 1/10/14/19 ("address space different than that of a second process"); anticipatory value: none.
  • US 4,811,393 → the keyed challenge/response concept in claim 1 and the "computed value vs. predetermined value" idea in claim 6; anticipatory value: none.
  • US 4,964,163 → challenge/response structure in claim 1 and the timeout limitation of claim 5; anticipatory value: none.
  • US 5,311,594 → fresh/random challenge per transaction (claim 1) and the "nonce of no persistent worth" concept of claim 4 by analogy; anticipatory value: none.
• If one were to rank relevance (not anticipation): US 4,964,163 (Motorola puzzle/answer + timeout) and US 4,811,393 (Bull keyed certificate exchange) are the two most relevant of record; US 4,521,846 and US 5,311,594 are the least. The strongest prior art actually raised against the patent was the IPR's Aucsmith + Chang combination (obviousness), which was never adjudicated because the IPR settled.

5. Confidence and open items (stated explicitly)

• High confidence: the four examiner-cited references, their titles, dates, assignees, and the fact that they are the only patent citations of record; the six family-cited references and their dates; the claim text of '550; the IPR number, petitioner, filing date, and settled outcome.
• Moderate confidence: that the IPR's "Chang" is EP 0 686 906 A2 / US 5,724,425 (from the earlier-generated section), and its correspondence to DE 69704684 T2.
• Low confidence / unverified: the exact identity and document number of "Aucsmith" (Ex. 1009); and a reference to "Berry" that appears in the petition excerpts ("the applicant argued that Berry disclosed broadcast of a puzzle to mobile receivers"). "Berry" is not among the four examiner-cited references, and I could not verify what it is. Treat both as open items requiring a PTAB/prosecution-history pull.
• Not legal advice. The above is a technical-art mapping, not a validity opinion or claim construction.

Primary sources: US5974550A (Google Patents, full text + citation lists) · US4811393A · US5311594A · uspto.report/US4521846 · Unified Patents US-4811393-A · IPR2017-01783 petition (USPTO PTACTS) · FreePatentsOnline US5974550

I'll verify the prior art references before building the obviousness analysis, since the analysis must be grounded in what these references actually disclose.

Now let me verify the additional references I need for a complete §103 picture — the Chang references and the two Intel applications incorporated by reference.

Obviousness Analysis — U.S. Patent No. 5,974,550 under 35 U.S.C. § 103

Scope and caveats. This is a technical obviousness analysis of the claims as I understand them from the authoritative text, using the prior-art references surfaced in the patent page's citations, family citations, and the spec's own incorporation-by-reference. It is not a validity opinion, claim construction, or legal advice. The '550 patent has an effective filing/priority date of 1997-12-12, so pre-AIA § 103 governs, and every reference must qualify as prior art as of that date.

1. The person having ordinary skill in the art (PHOSITA)

For the 1997 timeframe, a PHOSITA would be a software/systems engineer with a bachelor's degree (or equivalent) in computer science or electrical engineering and roughly 2–4 years' experience in computer security and cryptographic protocol design, including: symmetric/asymmetric cryptography and digital signatures; challenge–response authentication; and at least working familiarity with operating-system internals (virtual address spaces, cross-memory program calls) and software-distribution/piracy protection. This matters because § 103 is judged against that artisan, not a cryptographer alone or a kernel programmer alone — the invention sits precisely at the junction of those two fields.

2. The prior art, with prior-art qualification gating

2a. References cited on the face of the patent ("Citations (4)")

2b. Family-cited references ("Family Cites Families (6)") — status must be policed

Watch-out: three of the six "family cites" post-date the '550 priority date and cannot be § 102/§ 103 art. Any rejection built on them is defective. Only the first three are usable.

2c. References from the IPR petition (not on the face of the patent)

• US 5,724,425 (Chang, Sun Microsystems), filed 1994‑06‑10, issued 1998‑03‑03 — Method and apparatus for enhancing software security and distributing software. A compiler computes a message digest of the binary and encrypts it with the writer's private key to form a digital signature; the "software passport" (signature + license + binary) is distributed; the recipient platform verifies the digest against the signed value and refuses to execute altered code; "any unauthorized changes to the binary code … [are] evident through the comparison of the calculated and encrypted message digests." This is the "signed manifest" / verify-an-image-before-running teaching that maps directly to claim 6. European counterpart EP 0 686 906 A2. (Google Patents US5724425)
• US RE42,762 — relied on in the IPR. ⚠️ I could not verify this reissue's content before exhausting my search budget; treat its disclosure as unconfirmed.

2d. The references the patent itself incorporates by reference (the critical ones)

The specification states: "[d]etailed methods for creating the tamper resistant module and providing integrity verification processing with IVKs and bilateral authentication are disclosed in pending U.S. patent applications entitled 'Tamper Resistant Methods and Apparatus', Ser. No. 08/662,679, and 'Tamper Resistant Methods and Apparatus', Ser. No. 08/924,740, both … incorporated herein by reference."

These two Intel applications (the Aucsmith tamper-resistant-software work) supply everything the four face-cited references lack: self-decrypting, self-modifying, installation-unique, position-independent code; an Integrity Verification Kernel (IVK) that verifies an image against a supplied digital signature and fails completely if any byte is altered; a tamper-resistant compiler producing obfuscated images; and bilateral authentication between modules. ⚠️ I was unable to confirm their issued patent numbers within my search budget — I am giving the serial numbers exactly as the patent recites them (authoritative) and flagging the gap rather than guessing.

3. The § 103 gap: the face-cited art alone is insufficient

Standing on the four examiner-cited references alone, a § 103 rejection of claim 1 is weak. None of US 4,521,846, US 4,813,393, US 4,964,163, or US 5,311,594 discloses:

1. a module (as opposed to a data exchange) that is created by one process and sent to another;
2. that is tamper-resistant — resistant to observation/modification, self-decrypting, failing closed if altered;
3. that self-verifies the integrity of the host/executing process before releasing a secret; or
4. a secret released conditionally on that verification and then used as the key to answer the challenge.

That is the inventive core. Any credible § 103 case must therefore reach for Chang and/or the two incorporated Aucsmith applications to fill elements (2)–(4).

4. Element-by-element mapping of claim 1 (the strongest combination)

Ground 1 (primary): US 4,521,846 + US 5,724,425 (Chang) + Aucsmith '679/'740 + US 4,813,393 (Bull) + US 4,964,163 (Motorola)

Motivation to combine (why a PHOSITA would have done this)

1. Same field, same problem, shared motivation. All five references are in authentication/access control for a remote, untrusted counterparty. Bull, Motorola and Penzias authenticate the counterparty's credential; Chang authenticates the counterparty's code image. A PHOSITA facing the '550 background's stated problem — "general challenge-response protocols prove only that a verified channel between two endpoints sharing a secret has been set up, but one of the endpoints could be insecure" — would naturally want both, and the two techniques are complementary, not duplicative. KSR rationale: combining two known techniques to obtain the combined advantages each was known to provide.
2. The patent's own admission supplies the motivation and the mechanism. The specification concedes: "This can be done when the parties share the same address space by checking the contents of memory of the other party, computing its digital signature, and verifying its integrity." The applicant therefore admitted that signature-based integrity verification of a program image was known, and identified the only remaining obstacle as the address-space boundary. IBM '846 is the known answer to that obstacle. That converts a "novel architecture" argument into a "known technique, applied where it was known to be needed" argument — the KSR "predictable result" / "obvious to try" posture.
3. Explicit teaching in Aucsmith to run the module in a foreign address space. The tamper-resistant software is position-independent and requires no relocation, i.e., it is expressly designed to be executed somewhere other than where it was created — an express motivation to use it as the "agent sent to the remote process."
4. Motivation to avoid pre-shared high-value secrets. Bull's whole point is eliminating reliance on an identical pre-shared key; Motorola's puzzle is designed to defeat cloning. The '550 claim 1's use of a freshly minted secret in the module is the same design pull, so the combination meets the claimed "no previously communicated secret" advantage without hindsight.
5. Reasonable expectation of success. Each sub-step (send a program across address spaces; verify a digest against a signed value; use a keyed transform of a challenge; compare results) was independently known and mechanically combinable, with no apparent technical incompatibility. This is a predictable, not unpredictable, art.

Ground 2 (alternative, if one treats the tamper-resistant module itself as admitted art): IBM '846 + Chang '425 + Bull + Motorola + Penzias. Because the spec incorporates Aucsmith, the incorporated content is part of the '550 disclosure; a petitioner may argue the remaining elements are supplied by the incorporated material read with Chang. This is messier analytically and I flag it as the weaker framing.

5. Dependent claims

Claims 3–9 are the weakest link in the patent. Each maps to a discrete, express teaching in a single reference, and each is a conventional design choice. Claim 5 and claim 6 in particular read almost directly onto Motorola '163 and Chang '425 respectively.

6. Independent apparatus/medium claims (10, 14, 18, 19, 20–23)

• Claim 10 / claim 20 (challenger side; apparatus/MRM). Create the module+secret, create the challenge, send both, receive and decode the response. These are the method steps of claim 1 recast in apparatus form — obvious for the same reasons, since the references describe the systems (Bull's exploitation system; Motorola's central unit) performing exactly these acts. No separate § 101/112 analysis is attempted here.
• Claim 14 / claim 21 (responder side). Receive the module, execute it, recover the secret when integrity is verified, receive the challenge, encode with the secret, respond. Bull's target device (card) + Motorola's subscriber unit + Aucsmith's IVK module.
• Claim 18 / claim 22 (combined both roles in a single instruction set). Note — as flagged in the earlier section — this claim recites one program that both creates and executes the module. Obviousness here is aggravated: the '550 spec itself says that when the operating system and processor are the same (its example: WINDOWS NT/95 on PENTIUM®/PENTIUM® II), "a tamper resistant module created by one party will be a valid executable on the other party's system." That is an express admission that the same code base runs on both endpoints, supplying the motivation to implement both roles in one machine.
• Claim 19 / claim 23 (bilateral). Mutual authentication: each party sends a module + challenge and must answer correctly. Bull (both sides compute certificates), Motorola, and Aucsmith (express "bilateral authentication of partner modules") each point to symmetry; running the same protocol in both directions is the canonical, obvious extension of any one-way challenge–response protocol.

7. Where the patent owner has real defenses

A rigorous analysis must state the counter-arguments, not just the rejection theory:

1. The "module verifies the host" direction is not squarely taught anywhere. Chang verifies software about to be loaded/executed; Aucsmith verifies the module's own image and its partners'. The '550 requires the module to verify the integrity of the first process — the caller/host — and to withhold a secret until that succeeds. A patent owner will argue this is a genuine architectural difference and that no reference motivates checking the host from inside a dispatched guest module. That is the strongest non-obviousness argument, and it is reinforced by the '550 background's framing that a shared-address-space integrity check is possible but cross-address-space is not — i.e., a claimed "discovered problem."
2. § 103(c) may disqualify the two Aucsmith applications. They were Intel applications (Ser. Nos. 08/662,679 and 08/924,740), commonly owned with the '550 at the time the '550 invention was made. Pre-AIA § 103(c) prevents the use of commonly owned subject matter as § 103 prior art where that subject matter qualifies only under § 102(e)/(f)/(g). If Aucsmith qualifies only as a § 102(e) reference (U.S. filing before invention, publication after), it is removed as § 103 art — at a stroke deleting the only face-plausible source of the tamper-resistant/IVK element. A challenger would have to establish independent prior art (e.g., the Chang family, or the "Similar Documents" hits such as US 6,138,239 and US 7,073,062 — but note those post-date 1997 and do not qualify) for the same element. This is a substantial vulnerability in the obviousness case.
3. Secondary considerations — largely unproven. No evidence of unexpected results, long-felt need, or industry praise with a nexus has surfaced. The heavy citation of '550 by later Microsoft, Intertrust, NDS and Pacid patents (60 "cited by" entries) shows the document was influential, but being cited is not objective indicia of non-obviousness and carries no nexus. The 2017 IPR settled before institution (IPR2017-01783, terminated 2018‑01‑09), so there is no PTAB or court merits finding on validity one way or the other — a settlement is not probative of non-obviousness, and a failure of the petitioner to obtain institution is not an affirmance of validity.
4. "Different address space" is structural, not functional. IBM '846 shows the address-space machinery was entirely conventional by 1985; a claim that turns on that phrase alone adds little.

8. Bottom line

The most defensible § 103 case is: IBM '846 (different address spaces / cross-memory execution) + Chang '425 (verify a code image against a signed digest and fail closed) + the two incorporated Aucsmith applications (tamper-resistant, position-independent, self-decrypting module with IVK and bilateral authentication) in view of Bull '393 (secret-keyed challenge-response without a pre-shared per-device key) and Motorola '163 (puzzle/answer with timeout), further in view of Penzias '594 for claims 3, 4, 7–9. The motivation is a known problem admitted in the '550 background (challenge-response proves only a shared secret, not that the counterparty is unhacked) addressed by known, complementary techniques with a reasonable expectation of success. The decisive weakness of the case is that no single reference or articulated combination clearly teaches the module verifying the host process (rather than itself or the code being loaded), and pre-AIA § 103(c) may remove the Aucsmith applications as § 103 art by reason of common Intel ownership.

9. Explicit flags and contradictions

• Date inconsistency (not material to the analysis): the task header states the current date as April 26, 2026, while the patent-page fetch timestamp is 2026‑09‑29 and the earlier generated sections used both. I used the 1997‑12‑12 priority date throughout; the discrepancy does not affect the § 103 reasoning.
• Numbering check — no conflict found: the earlier "Patent summary" cited the Chang reference as "US 5,724,425"; my search resolved it to US 5,724,425 (app. 08/258,244, Sun Microsystems, 1994‑06‑10 → 1998‑03‑03). The two strings are identical (5,724,425), so there is no transposition error to report — contrary to what one might expect given the similar-looking digits.
• Unverified despite searching: US RE42,762 and EP 0 686 906 A2 (content), and the issued patent numbers corresponding to Intel Ser. Nos. 08/662,679 and 08/924,740. I exhausted my search budget and am not guessing these identifiers. The serial numbers are quoted exactly as the authoritative patent text recites them.
• Cannot confirm: the three "Family Cites Families" references with 1998 priority dates (US 6,263,437 B1, US 6,918,035 B1, KR 100285791B1) are not prior art to '550 and should not be cited against it.
• I express no conclusion on validity, enforceability, infringement, or claim construction. Claim 18/22's single-instruction-set-both-roles structure in particular would require the prosecution history to interpret reliably.