A declassified internal CIA memorandum that briefly outlines how Jim Sanborn was contracted to create the Kryptos installation.

According to most sources, including the CIA, the Kryptos sculpture was privately dedicated on November 3, 1990. However, a transcript of a speech given by then CIA Director William Webster during the ceremony is dated November 5, 1990. In an interview with CBC podcast host Jian Ghomeshi recorded in late 2010, sculptor Jim Sanborn stated the dedication ceremony was held on November 5, 1990. In an interview with CNN a few days later, Sanborn told them the dedication ceremony was held on November 6, 1990. The source of the November 3 date is unclear, and the November 6 date was likely an error, so November 5 seems like the most probable date of the Kryptos dedication ceremony.

This is an Associated Press article that was published in many different newspapers around the same time. See below for a non-paywalled online reprint by the Los Angeles Times.

The word "κρῠπτός" is an ancient Greek word for "hidden" or "secret" which can be transliterated into "kryptos". The suffix crypto- in words like cryptology, cryptography, and cryptanalysis stem from the same root word.

Jim Gillogly first revealed his partial decipherment of Kryptos on June 8, 1999, in a message posted to the cryptography-oriented Usenet group sci.crypt. He went on to explain that it took him "four evenings" to solve K1-K3 using his own custom-made software running on a Linux computer with an Intel Pentium II 400 MHz CPU. Upon calling the CIA's Office of Public Affairs, Gillogly was informed that CIA physicist and technical analyst David Stein (see reference #13) and an anonymous group of NSA cryptanalysts (see reference #14) had already solved K1-K3.

CIA physicist and technical analyst David Stein started working on Kryptos around 1991 and solved K2 on February 21, 1998, followed shortly by K3 and K1 (in that order). According to Stein, he spent approximately 400 hours across seven years working on Kryptos and used nothing but pencil and paper. Stein held a briefing about his partial cracking of Kryptos inside a large auditorium at CIA headquarters on March 26, 1998, which attracted roughly 250 attendees, including CIA cryptographer Edward Scheidt.

Shortly after Jim Gillogly announced that he'd solved K1-K3 in 1999, the National Security Agency reported that three unnamed NSA cryptanalysts had solved the same segments at some undisclosed time in the past. In 2000, it was revealed the team was actually composed of four cryptanalysts, including Ken Miller and Dennis McDaniels, who partially decrypted Kryptos in late 1992. Following a series of Freedom of Information Act requests by multiple individuals, internal NSA documents outlining their efforts to solve Kryptos were released in 2013 and 2014. Shortly thereafter, the NSA made the documents available on their website. The documents show there were two separate attempts by the NSA to solve Kryptos. The first attempt was made by a small group of NSA cryptanalyst interns following a trip to CIA headquarters in late 1991. They formed what they called the CIA Sculpture Study Group (CSSGp) and held a few meetings, but never cracked the code. In late 1992, during an internal NSA awards ceremony known as the Gold Bug, then NSA Director John McConnell relayed a challenge issued by then CIA Deputy Director and former NSA Director William Studeman for the NSA to solve Kryptos. The aforementioned four-member team accepted the challenge, and three of them solved one segment each. First K2, then K3, followed by K1. According to one document, these segments were solved "within two days" of accepting the challenge, and they spent "another day" on K4 before giving up. However, according to another document and the 2000 press article, K1-K3 were deciphered "within one month", and they spent "another month" on K4. Nevertheless, the NSA's efforts were relayed to Studeman on June 9, 1993, in an internal memo, but it doesn't seem like it became general knowledge within the CIA until the late 1990s, when David Stein (see reference #13) and Jim Gillogly (see reference #12) independently solved K1-K3.

General information about frequency analysis.

Frequency tables of the English language differ depending on sample size and the type of text analyzed. The table referenced here is the same source Wikipedia uses for its article about letter frequency.

I only used the first three sentences from this article and ignored the text within parentheses.

A brief summary of the difference between the two major cipher classes, transposition and substitution.

I was unable to find a direct quote attributed to Sanborn by a reputable source where he explicitly says one or all of the Kryptos typos were deliberate. In some articles, it's reported that that's what Sanborn said or has said in the past without a direct quote. In one article from 2020, Sanborn was quoted as saying the IQLUSION typo was intended "to mix it up." In an interview from 2005, Sanborn described the same typo as "a clue." However, speaking generally about all the typos, he said that he was "sloppy" and "didn't take tremendous care" when he wrote the plaintext before ambiguously conceding, "if you want to say it's intentional, it's intentional." Behind closed doors, Sanborn has apparently confirmed the IQLUSION and UNDERGRUUND typos were deliberate, according to Kryptos enthusiast Elonka Dunin and others. He has also implied the DIGETAL typo in the Morse code was deliberate. But he has apparently declined to answer whether DESPARATLY was intentional.

When properly formatted, the coordinates in the K2 plaintext looks like this: 38°57'6.5"N 77°8'44"W.

Between August 23-25, 2013, the American Cryptogram Association hosted its annual convention in Norfolk, Virginia. One of the speakers at the convention was retired CIA cryptographer Edward Scheidt who was specifically there to talk about Kryptos. From what I can tell, Jim Sanborn was not a scheduled speaker, but later joined Scheidt to answer some questions about Kryptos during a post-convention dinner. Below is a link to a page where an anonymous Kryptos enthusiast has quoted and summarized some of the questions and answers. Sanborn apparently confirmed something was buried at CIA headquarters, as indicated by the K2 plaintext. The object was a bronze geodetic survey marker by the United States Geological Survey, which had "little cross hairs on the top of it." But when Sanborn and his wife Jae visited CIA headquarters "a couple of years" before 2013, the marker had been removed. As far as I know, this is the only time Sanborn has shared this information. Sanborn also implied there was yet another segment after K4, which he described as K5. But he's probably referring to the puzzle within the plaintext, which he has previously described as "a riddle within a riddle." Sanborn has also described K4 as the final segment on multiple occasions, and K4 is already very short at roughly 97 characters, making it unlikely to be a two-parter.

This is a page operated by a Kryptos enthusiast named John B. Wilson. Under the heading "from Elonka Dunin's 2003 roadtrip", Wilson has summarized an interview with Sanborn conducted by fellow Kryptos enthusiast Elonka Dunin in 2003. As far as I can tell, this information is not available anywhere else. I assume it was initially shared on Dunin's Kryptos community on Yahoo Groups back in the early 2000s, which was only accessible to registered members before the service was discontinued in 2020. Apart from the Dunin interview, there are many other informative quotes and summaries on this page.

According to the standardized style guide of the American Radio Relay League, the letter X is "used to denote a period" in radiograms.

While discussing the IDBYROWS/XLAYERTWO mishap, Jim Sanborn explained that "because a period isn't really easy to encode, I chose to use Xs."

When asked, "Who do you think is most likely to solve K4?" in this email interview, Sanborn's tongue-in-cheek response was, "A Vulcan."

This is an archived recording of an episode of a CBC Radio podcast known as Q which, at the time, was hosted by Jian Ghomeshi. It includes a segment about Kryptos towards the end, during which Jim Sanborn is interviewed. The episode aired just three days after Sanborn disclosed the first K4 clue. When asked about author Dan Brown incorporating Kryptos into his fiction novels The Da Vinci Code and The Lost Symbol, Sanborn said he considers art and popular culture to be two completely separate worlds. He described himself as a "non-fiction artist" while describing Brown's work as "pseudo-science" and did not seem to appreciate being associated with the world of pop culture.

CIA cryptographer Edward Scheidt assisted Jim Sanborn with the creation of the Kryptos code. It's not entirely clear what Scheidt does or does not know, but, generally speaking, it sounds like Scheidt taught Sanborn how to encrypt and decrypt messages using a bunch of cryptographic techniques. Sanborn then secretly modified those techniques to prevent Scheidt from knowing exactly what he did.

General information about crib dragging.

According to Kryptos enthusiast Patrick Foster, he was the one to discover that the keywords PALIMPSEST and ABSCISSA could be combined to form the anagram PS ITS AS SIMPLE AS ABC on November 10, 2003.

In this article, Kryptos enthusiast Monet Friedrich presented a theory for how to acquire the K1 keyword PALIMPSEST using palindromes within the Morse code. The idea is to extract some of the palindromes and stack them on top of each other so that PALIMPSEST is spelled out vertically in the central column. But this theory is far from perfect. First of all, all the palindromes created by the many repeating Es must be ignored, which leaves us with eleven palindromes. They are RPR, TAT, LL, TIT, ERPRE, ESE, MEM, ISI, ITI, SOS, and TERPRET. Note that while the repeating E palindromes must be ignored, the theory also requires the double L palindrome to be included, even though it's also composed of a single repeating letter. Anyway, if you look at the center of the eleven palindromes listed above, you'll find the letters P, A, L, I, P, S, E, S, T, O, and P. For the theory to work, we must ignore the O from SOS and one of the three Ps. We must also overlook the fact there's no central letter in the two-letter LL palindrome. At this point, we have all the letters for spelling PALIMPSEST except the M in the middle. To acquire an M we need to combine the two Morse code fragments LUCIDEE and MEMORYE into LUCIDEEMEMORYE which gives us the letter M via the palindrome EME. We can now finally spell out PALIMPSEST using the extracted palindromes. Without making these rather arbitrary concessions, the theory falls apart. That being said, it's still interesting that almost all the letters of PALIMPSEST can be acquired in this way, so even if Friedrich's theory is flawed, she might've been on the right track.

The are many ways to decrypt K3. I've referenced a few popular examples below.

Jim Sanborn spoke at a conference in 2013 known as Big Tech Day 6 and showed some slides during the presentation. One of the slides (numbered 1371220840) was a previously unreleased image of the encoding chart Sanborn used to create the K3 ciphertext.

It's unknown whether the final question mark belong to K3 or K4. Jim Sanborn has occasionally described K4 as being 97 character long, suggesting the question mark belongs to K4. But he has also given more ambiguous answers so no one really knows.

When the NSA partially deciphered Kryptos back in 1992 (see reference #14), they were confused by the sentence ID BY ROWS at the very end of K2. This confusion persisted after David Stein (see reference #13) and Jim Gillogly (see reference #12) independently solved K2 in the late 1990s. On October 11, 2005, Kryptos enthusiast Monet Friedrich was crib-dragging different keywords through K2 when she noticed the keyword BSCISSAA would produce PLAYER TWO instead of ID BY ROWS. Barely six months later, on April 19, 2006, Jim Sanborn called Kryptos enthusiast Elonka Dunin to inform her that he'd made a mistake by excluding an S between the ciphertext letters E and W near the end of K2. So the final sentence of K2 was meant to read X LAYER TWO rather than ID BY ROWS. The fact that this omission still produced legible text was pure coincidence. Typically, such omissions will disrupt the cipher and generate gibberish. Sanborn did not realize his mistake earlier because he thought ID BY ROWS was some sort of "cryptographer's jargon".

While Jim Sanborn has ensured there's nothing wrong with the K4 plaintext, it's less clear whether he's gone back and double-checked the intended step-by-step process for extracting said plaintext.

Jim Sanborn disclosed the K4 BERLIN clue in this article.

Jim Sanborn disclosed the K4 CLOCK clue in this article.

In this article, Jim Sanborn hints that the phrase BERLIN CLOCK in the K4 plaintext is a reference to the Mengenlehreuhr (Set Theory Clock) in Berlin which is also known as the Berlin Clock.

Jim Sanborn disclosed the K4 NORTHEAST clue in this article.

In the midst of the COVID pandemic, Jim Sanborn privately disclosed the EAST clue to several people, beginning in April 2020. Sanborn later explained that he was "a tad bored" and wanted to see "how long it would take to flow through the community". Initially, when people tried to inform the broader Kryptos community, they were met with skepticism. But in August 2020, The New York Times journalist John Schwartz contacted Sanborn who confirmed the clue was genuine.

Jim Sanborn spoke at a conference in 2013 known as Big Tech Day 6 where he described himself as an "anathemath", a nonce word meaning a person who dislikes mathematics. See reference #41 for more information about this reference.

In this interview, Edward Scheidt said he "masked the English language" in K4 to make it "more of a challenge" than K1-K3. The plaintext of all four segments is written in English, but Scheidt deliberately tried to "disguise" the English language in K4 to disrupt the use of "frequency counting". There are countless ways of obfuscating the language of the plaintext, but cipher stacking is at least consistent with these comments.

The authors of this paper explored the possibility that K4 was encrypted using the Hill cipher, invented by American mathematician Lester Hill in 1929.

Should Jim Sanborn pass away before K4 is deciphered, the solution to K4 will most likely be passed on to someone else. It has apparently been stashed away in one or more secure locations, known only to Sanborn and a select few undisclosed confidants. Should the need arise, they'll be able to access the K4 solution and either reveal it to the world or keep the mystery going. Sanborn has also toyed with the idea of putting the solution up for auction.