# Released clues and their limits

Research cutoff: 14 September 2026. This register separates fixed plaintext constraints, attributed statements, historical context, and proposed interpretations. No newly authenticated positional plaintext letters were found beyond the existing 24. Positions below are one-based, inclusive.

| Disclosure | Fixed observation | Consequence |
|---|---|---|
| November 2010: BERLIN | 64–69: NYPVTT → BERLIN | Six fixed plaintext letters. |
| November 2014: CLOCK | 70–74: MZFPK → CLOCK | Five additional letters, adjacent to BERLIN. |
| January 2020: NORTHEAST | 26–34: QQPRNGKSS → NORTHEAST | Nine fixed letters. |
| April 2020 private circulation; August public release: EAST | 22–25: FLRV → EAST | Four more letters, preceding NORTHEAST. |

The clue chronology and placement are recorded by [Elonka Dunin](https://www.elonka.com/kryptos/); contemporary CLOCK reporting is in [Kim Zetter's November 2014 article](https://www.wired.com/2014/11/second-kryptos-clue/). Treat the contiguous strings as plaintext constraints. Their role as compass instructions, keys, or routes remains a hypothesis. EAST was circulated before it became widely public; publication month alone is not its first disclosure date.

In Dan Vergano's [12 November 2025 report from Sanborn's Spy Museum presentation](https://www.scientificamerican.com/article/cia-kryptos-puzzle-creator-releases-final-clues/), the artist identifies the World Clock as the referent of BERLINCLOCK and links the solution to his 1986 Egypt trip and the fall of the Berlin Wall. The report also describes message delivery as the work's theme, and K5 as a future 97-character message with some shared coded words/positions and a connection to K2's buried-object theme. These are reported author disclosures, not a published encryption recipe. The linked [open-letter scan](https://www.documentcloud.org/documents/26229389-adobe-scan-nov-12-2025/) could not be retrieved in this pass; its exact wording has not been independently read. The article itself contains a 79-versus-97 typo, so it is not transcription authority.

The World Clock has 24 sectors, a rotating hour ring, city labels, and a compass-rose mosaic beneath it. Berlin's [official description, updated June 2023](https://www.berlin.de/sehenswuerdigkeiten/3561749-3558930-weltzeituhr.html), also records subsequent corrections to city time zones and additions of cities. A modern list is therefore not automatically a valid 1988–1990 key. None of these physical facts establishes which feature, if any, encodes K4. Earlier set-theory-clock interpretations should be deprioritized as explanations of the named clock; their negative tests remain valid only for their actual models.

Two historical source issues matter. Sanborn's [2009 oral-history transcript](https://www.aaa.si.edu/download_pdf_transcript/ajax?record_id=edanmdm-AAADCD_oh_283557), conducted by Avis Berman, recalls Egypt visits in 1977 and 1984, whereas the 2025 clue reporting specifies 1986. Preserve that discrepancy rather than silently changing a date. The oral history describes reflected light, tombs, museum displays, and Abu Simbel as artistic influences; none is an authenticated K4 key. Similarly, 2025 reports say the archive was sealed, while the Smithsonian's [current collection-use page](https://www.aaa.si.edu/collections/jim-sanborn-papers-22298/how-to-use) describes the collection as open for research by appointment. The specific K4 materials' access status is unresolved; the generic finding aid does not establish that they are available online.

Sanborn's [January 2005 interview with Kim Zetter](https://www.wired.com/2005/01/questions-for-kryptos-creator/) supports three useful limits: visiting CIA grounds is unnecessary for decipherment; he modified systems and used visual approaches; earlier text contains instructions referring to later text. It does not authenticate any particular proposed use of K1–K3. The claim that K4 must be solved independently of all earlier plaintext remains insufficiently sourced in the reviewed secondary references and should not eliminate such hypotheses.

The [1991 Sanborn interview transcription](https://www.thekryptosproject.com/media/transcripts/pdfs/1991a.pdf), prepared in 2004 by Dunin and Andy Zimolzak, distinguishes readable English from interpreting the artwork's deeper meaning. The [1991 Scheidt transcription](https://www.thekryptosproject.com/media/transcripts/pdfs/1991c.pdf) emphasizes the undisclosed keys or method and the paucity of comparable messages. Neither gives a finite search bound or promises an easy mathematical solution.

Current stewardship is no longer accurately described as merely an anonymous buyer. [Paradigm's own 2026 announcement and video transcript](https://www.linkedin.com/posts/paradigm-xyz_97-characters-35-years-unsolved-one-sculpture-activity-7471199131445772289-fzBf) identifies the firm as steward and describes an answer verifier designed not to expose the plaintext to its operators. Its ten new CTF puzzles are separate training challenges. Sanborn's remarks in the video about numbers and creative systems do not add a positional crib or establish a specific algorithm. The primary Paradigm website was inaccessible through the research tool; no verifier was called and no submission fee was paid.

The [RR Auction lot](https://www.rrauction.com/auctions/lot-detail/350761607302001-the-complete-secrets-of-kryptos-jim-sanborns-private-archive/) distinguishes plaintext, coding system, and alternate K5 materials, while explicitly saying the auctioneer had not independently inspected the secret materials. Archival discovery of plaintext is not itself a public derivation of the encryption. Likewise, a proposed method fitted to an alleged plaintext is not authenticated by the existence of the real archived plaintext.

Admission rule: use the 24 established letters as exact constraints. Use historical and thematic clues to select a small independently defined mechanism. Do not convert interpretive remarks, ambiguous dates, proposed plaintexts, video graphics, or a training-puzzle solution into additional hard constraints.