What does a public statement about a known criminal owe to the documentary record? What is its half-life? When the records eventually arrive — through a Department of Justice release, a court disclosure, an estate filing — who carries the responsibility to update what was said: the speaker, or the public?

These are not new questions in journalism. They are unusually pointed in this case.

By the time Neri Oxman first arranged to sit down with Jeffrey Epstein in October 2015, he had been a registered sex offender for seven years. His 2008 Florida plea covered procuring a minor for prostitution. The non-prosecution agreement had already become a national scandal. The names of his victims were already on the public record. None of this was obscure or hidden in 2015.

By the time her husband, Bill Ackman, described their family's contact with Epstein in a long X thread in January 2024 as a single "platonic encounter," Epstein had been dead for four years and the documents from his estate had begun to surface in federal disclosures. Some of those documents are now public. They are the basis of this piece.

On 30 January 2026, the U.S. Department of Justice released the first tranche of materials under the Epstein Files Transparency Act. The release contained more than 33,000 pages of correspondence and scheduling documents from Epstein's estate, organized into twelve Data Sets accessible at justice.gov/epstein. A search of those files for "oxman" and "neri" returns more than four hundred unique document IDs spanning the period August 2015 to January 2019.

This article works through what those documents show, what was said publicly about the relationship in 2019, and what was elaborated about it in 2024. The question it asks is the one in the opening paragraph. It does not allege any criminal conduct on Neri Oxman's part. It does not allege complicity in Epstein's crimes. It documents a contradiction between public statements and primary-source correspondence, and it leaves the conclusions to the reader.

The 2019 statement

On 13 September 2019, two months after Epstein died in federal custody awaiting trial on sex trafficking charges, and as the institutional reckoning at MIT was reaching its peak, Neri Oxman published a statement on Medium that was reproduced in the Boston Globe. The statement acknowledged a $125,000 donation from Epstein to her research group, the Mediated Matter Group, in 2015. It described her introduction to Epstein as having taken place at the MIT Media Lab through Joi Ito, then the Lab's director. It characterised her contact with Epstein as a single in-person meeting in October 2015. It expressed regret for not having researched Epstein's background before that meeting.

The phrase that carried the most editorial weight in the statement, and that has been quoted by every subsequent journalist who has written about it, was the description of the October 2015 meeting as the "first and only time" Oxman met Epstein in person.

The statement also acknowledged that Epstein had subsequently visited the Media Lab on other occasions, that Oxman had been aware of these visits, and that she had declined to attend them. It described correspondence with Epstein that followed the October 2015 meeting as limited and professional in nature.

Two further details from the statement bear on what follows. First: the donation was acknowledged as $125,000 to her research group, plus a separate $125,000 to the Media Lab itself. Second: the statement characterised Epstein's contact with Oxman's studio as a single indirect visit by way of Joi Ito, and asserted that Oxman had not sought his patronage.

The statement was widely treated by the press in 2019 as a candid acknowledgment by a senior MIT figure of an embarrassing connection. It was not initially treated as a contestable factual account. The documents released in January 2026 invite that re-examination.

August 2015 — before the introduction

The earliest dated document in the released set referring to Oxman is not from October 2015. It is from 23 August 2015, two months earlier. The document is an email from Joi Ito — then the director of the MIT Media Lab — to Jeffrey Epstein. It refers to a Krebs cycle image that Oxman had drawn, and notes that Epstein had already seen it.

The document is filed under DOJ EFTA02709904. It places the active pre-marketing of Oxman's work to Epstein, by the director of the MIT Media Lab, two months before the meeting that Oxman would publicly describe four years later as the start of her contact with Epstein. The framing is not that Ito was introducing Oxman's work to Epstein for the first time; the framing is that Epstein was already familiar with it. The relationship between Joi Ito and Epstein at this point was, on the documentary record, sufficiently developed for them to be casually exchanging references to MIT faculty research.

This does not contradict Oxman's 2019 statement directly. The statement said the meeting in October was the first; it did not say nothing of her work had been seen by Epstein before. But it does materially complicate the account that has dominated subsequent press coverage — that she was, in essence, an unknowing subject brought into Epstein's orbit through an institutional misjudgment by Joi Ito. The August 2015 document indicates that Ito's introduction of Oxman to Epstein in October was the culmination of a longer cultivation, not its beginning.

October 2015 — the meeting that wasn't at MIT

The October 2015 meeting that Oxman's 2019 statement described as the sole occasion of her direct contact with Epstein has been the subject of substantial press attention. It has been variously described as taking place at the MIT Media Lab on a Saturday afternoon, with Joi Ito as the principal facilitator. The 2019 Boston Globe coverage, which carried Oxman's statement, treated it as such.

The DOJ documents released in January 2026 contain a detailed scheduling thread for an October 2015 Cambridge meeting between Epstein and Oxman. They show that the meeting was planned for Friday 16 October 2015. They show that the location was Martin Nowak's Institute, 1 Brattle Square, Suite 6 — that is, Harvard's Program for Evolutionary Dynamics, an Epstein-funded research centre run by a professor whom Harvard would later discipline for those very ties. They show that Oxman cancelled the meeting, on her own initiative, the afternoon before it was scheduled to occur.

The relevant chain begins with EFTA00338218, dated 9 October 2015. Lesley Groff, Epstein's personal assistant, writes to Oxman to confirm the proposed time and location. Oxman, in EFTA00338206, replies the same day to confirm her availability. Over the following six days, Lesley Groff sends multiple follow-ups asking for confirmation; on 15 October, in EFTA00337406, she writes to Joi Ito asking for Oxman's mobile number because she has not been able to reach her directly. On the afternoon of 15 October 2015, Kelly Donovan — Oxman's executive assistant at the MIT Media Lab — sends a formal cancellation to Lesley Groff (EFTA00337508). Oxman herself, in EFTA00337514, sends a personal cancellation a few minutes later: she apologises that she will not be able to make it, and writes that she looks forward to meeting on a future occasion.

The full schedule for Epstein's planned 16 October 2015 Cambridge day is preserved in EFTA00337457 and EFTA00337483. It places the day at Martin Nowak's institute, with the following appointments: Joscha Bach and an unspecified group of AI colleagues at 11:00am; Oxman at 12:30pm (cancelled); Jeremy Rubin and a cohort of MIT cryptography students at 1:30pm — the named MIT students who were to attend included Sunoo Park, Ranjit Kumaressan, Alex Chernyakhosvsky, and Ravi Bajaj; Marvin Minsky and his wife Gloria Rudisch at 4:00pm; Martin Nowak alone at 5:00pm; and dinner at the Institute with Nowak and Noam Chomsky at 6:00pm.

This is the documented record of what was planned for 16 October 2015. The meeting Oxman publicly described in 2019 — first and only, at the MIT Media Lab, with Joi Ito as facilitator — is not visible in the DOJ release for that month. There are several possibilities. There may have been a separate, undocumented meeting at the Media Lab on a different October 2015 date that the search terms used did not surface. The 2019 account may have collapsed multiple events into one. The location of the original meeting, as remembered four years later in a public statement, may have been incorrect.

What the documents foreclose is the possibility that the 16 October scheduling thread is the meeting Oxman described. It is not. It is a meeting at Harvard, at Epstein's invitation, that Oxman cancelled.

A separate document filed under EFTA00843481, dated the same evening, contains a brief exchange within Epstein's circle in which his office tracks the cancellation count. The exchange notes that Oxman has cancelled three times — corrected within the thread to twice — and a TED talk link is shared. The document indicates that Oxman's prior cancellations had been tracked internally by Epstein's office well before the October 2015 meeting was rescheduled. By 16 October 2015 the relationship between Oxman's office and Epstein's was sufficiently developed for cancellation patterns to be a matter of contemporaneous observation.

The Harvard thread

The Martin Nowak connection is, on its own, a separate story. Martin Nowak ran Harvard's Program for Evolutionary Dynamics. Harvard's internal review of his Epstein ties resulted in his being barred from advising graduate students or pursuing new research projects in 2020. His institute received an estimated $9.4 million from Epstein-affiliated foundations and individuals between 2003 and 2007.

What the DOJ documents add to that story is the place his institute occupied in Epstein's wider Cambridge network. On the planned 16 October 2015 day alone, Nowak's institute was scheduled to host meetings with Oxman, four MIT cryptography students, Joscha Bach, Marvin Minsky and his wife, and Noam Chomsky over dinner. This is a documentary record of a Cambridge cohort being assembled by Epstein's office and Nowak's institute, across both Harvard and MIT campuses, on a single afternoon. None of these individuals — with the exception of Nowak himself and the named students — has been publicly connected, before this release, to that specific gathering.

Marvin Minsky's connection bears particular note. Minsky was the original 2002 MIT recipient of Epstein money — a $100,000 unrestricted gift documented in the 2019–2020 Goodwin Procter review of Epstein-MIT financial ties. Minsky died in January 2016, three months after the planned 16 October 2015 meeting. The documents place his last documented meeting with Epstein, attended with his wife, in October 2015 at Nowak's institute. Gloria Rudisch — Minsky's wife — appears again, separately, in a 2019 document discussed below.

A dedicated piece, The Harvard meetings, works through this thread in detail.

November 2015 — the apartment, the dinner, the proactive request

Three weeks after Oxman cancelled the 16 October Cambridge meeting, on 16 November 2015, Lesley Groff again wrote to Oxman with an invitation. The terms had escalated. Epstein was hosting Woody Allen and his wife Soon-Yi Previn for dinner at his Manhattan home on 18 November. He wanted Oxman to attend.

The same day — 16 November 2015 — Epstein himself wrote to Lesley Groff. The email is filed under EFTA02067746. Subject line, garbled by the OCR: "Re: [Neri] Oxman." The body of the message is three words: she can have apt.

The phrase corroborates a detail first reported by the Boston Globe in January 2026 — that Epstein had offered to put Oxman up in one of his Manhattan apartments. The documents released in 2026 contain the primary-source authorisation for that offer, in Epstein's own words.

What follows over the next 48 hours is documented in EFTA00335379, EFTA02067676, and EFTA02067702. Allen and Previn cancel the Wednesday dinner. Lesley Groff offers lunch instead — "Wed, Thurs or Fri would work." Oxman replies the following morning, on 17 November, asking whether Epstein might be available to see her on December 5 in New York instead. Lesley Groff offers December 4 in Cambridge as an alternative. Oxman accepts the 4 December morning meeting. The 10:30am meeting is confirmed by Kelly Donovan in EFTA02067927.

That meeting did not take place either. On 19 November 2015, in EFTA02067881, Lesley Groff wrote to Kelly Donovan to say that Epstein's schedule had changed — he would be in Cambridge on Saturday 5 December rather than Friday 4 December. Donovan replied that Oxman would not be back in Cambridge on the 5th, and removed the hold on her calendar. Lesley Groff's response: "it's a shame we are having such a difficult time getting them together!"

Three documented planned meetings in the October-December 2015 window. All three failed: one cancelled by Oxman, one cancelled when Allen and Previn pulled out and the rescheduled Cambridge alternative also fell through. Across the same window, Epstein authorised the use of his Manhattan apartment by Oxman, in writing.

The bearing of this on Bill Ackman's 2024 framing — his characterisation of his wife as not having accepted any invitation from Epstein — will be discussed below. For the moment the relevant point is narrower. The 2019 public statement described one in-person October 2015 meeting at the Media Lab as the entirety of Oxman's contact with Epstein. The documents released in 2026 show three planned meetings in the period October-December 2015, none of which is at the Media Lab and none of which appears, in the released set, actually to have taken place. The first-and-only meeting, as publicly described, is not visible in the DOJ documents for that period.

April 2016 — the fourth invitation

On 31 March 2016, six months after the 16 October meeting Oxman had cancelled, Lesley Groff again wrote to Oxman. The email, EFTA00326938, is short. Epstein would be at the Media Lab the following day; would Oxman be available to stop by?

The full schedule for Epstein's 1 April 2016 Media Lab visit is preserved in EFTA00326947 and EFTA00327006. It places his day at the Media Lab in Room 245. The schedule lists meetings with Seth Lloyd at 10:30am, Martin Nowak at 11:30am, Joscha Bach informally throughout the day, Joe Jacobson at 12:30pm, Jeremy Rubin and others at 1:30pm, and Joi Ito at 3:00pm. Larry Summers is on the schedule for dinner at a Boston restaurant at 7:00pm. Oxman's status is listed as: "TBD Neri Oxman... WAITING FOR REPLY."

Kelly Donovan responded on Oxman's behalf later the same morning, in EFTA00327028. The reply was warm in register: while Neri would love to meet with Jeffrey, she was returning from travel and would not be available. The phrasing distinguishes between intent and circumstance.

This is the fourth documented invitation by Epstein's office to Oxman in the period October 2015 to April 2016. It is the first in the released set actually to have been at the MIT Media Lab. It was declined, but not in terms suggesting any reluctance about future contact.

2017 — funding, gratitude, gifts

The DOJ release contains a 2017 correspondence thread between Oxman and Epstein, filed under EFTA009588. The thread comprises ten documented exchanges across June, August, and November 2017, of which seven are written by Oxman. They concern, in sequence: a June 2017 thank-you note for prior support of the Vespers project, with offered photos of the work; an August 2017 visit to New York with a request to meet; an August 2017 funding ask of $180,000 for a new death-mask collection (with a $50,000 contingency for additional portraits); a November 2017 thread continuing the same correspondence and including an unspecified gift offer.

The November 2017 thread, EFTA009588, carries a subject line written by Oxman in transliterated Hebrew: From Neri, toda ainsofit — a phrase rendering roughly as a superlative of thanks. Within the thread Oxman writes to Epstein on 9 November 2017 that her ears had been burning the previous weekend as she and Joi Ito had spoken of him and praised him. She offers him an unspecified small gift and asks where it should be sent. Epstein's reply provides his Manhattan home address.

The August 2017 thread contains the funding request and the line that gives the dedicated piece on this subject its title. Writing on 10 August 2017 in connection with a planned next collection, Oxman tells Epstein: "Vespers would have not come to life without you." A subsequent message in the thread, on 14 August, sets out the budget — $180,000 for a remake of the Death Masks series, with an additional $50,000 contingency. She invites Epstein to attend the SFMOMA / MoMA retrospective scheduled for 2020. A separate June 2017 message, also within EFTA009588, recounts the completion of the death-mask series for the London Design Museum and notes that the work would not have been completed without his support.

The 2019 Boston Globe statement disclosed the $125,000 donation in 2015 but did not mention the 2017 correspondence, the additional funding asks, or the gift sequence. The dedicated piece Vespers and the funding case works through the 2017 thread in detail. The relevant point for the lead is narrower: the 2019 framing of the relationship as having effectively ended after the October 2015 meeting cannot be reconciled with what Oxman herself wrote, and what she signed her name to, in late 2017 — over nine years after Epstein's 2008 conviction as a registered sex offender.

January 2019 — the Brockman event

The most recent dated document in this set referring to Oxman is from 4 January 2019. It is filed under EFTA01009359. It is not from Oxman, and it is not addressed to her. It is from John Brockman of the Edge Foundation, addressed to Epstein, and it concerns an upcoming Cambridge intellectual event scheduled for 21 February 2019.

The event, branded after Brockman's then-forthcoming book on artificial intelligence and consciousness, was to be held at the Brattle Theatre in Cambridge with a dinner at the Charles Hotel for sixty guests. Brockman lists fifteen confirmed speakers. Among them: Mary Catherine Bateson, George Church, Kate Darling, George Dyson, Peter Galison, Neil Gershenfeld, Jennifer Jacquet, Caroline Jones, David Kaiser, Seth Lloyd, Neri Oxman, Sandy Pentland, Steven Pinker, Gloria Rudisch, and Stephen Wolfram.

Epstein RSVP'd in two words to confirm his attendance. Brockman replied on the same chain warning him not to show up: "it will end in tears, mine."

This document is the most recent in the released set linking Oxman, by speaker billing, to the same Cambridge intellectual cohort whose senior figures — Pinker, Lloyd, Pentland, Church — have been variously and separately documented as having had their own contact with Epstein. The presence of Gloria Rudisch — Marvin Minsky's widow — on the same speaker list, more than three years after she and Minsky had attended the planned Nowak Institute meeting with Epstein in October 2015, draws an additional line through.

What the document does not establish is whether Oxman knew Epstein had been invited to the event, or whether she would have spoken at it had he attended. What it does establish is that as of 4 January 2019 — six months before Epstein's arrest, four years after the relationship was supposed to have effectively ended — she remained in the same Cambridge institutional circle through which the original 2015 introduction had been made.

Bill Ackman, January 2024

On 8 January 2024, Bill Ackman — Oxman's husband, founder of Pershing Square Capital Management, and one of the most prominent activist investors on Wall Street — published an extended statement on X. The thread was prompted by a separate professional matter and was framed as a defence of his wife against renewed media interest in her Epstein connections. It ran to several thousand words.

The thread made specific factual claims that are testable against the documents released two years later. They include:

That Oxman met Epstein in person on a single occasion. On the documentary record discussed above: at least four invitations were issued in the period October 2015 to April 2016, three of which were planned in detail. There is no document in the released set that places Oxman in a meeting with Epstein in October 2015 at the Media Lab.

That Oxman never accepted an invitation from Epstein. On the documentary record: Oxman accepted the 18 November 2015 dinner invitation. When that dinner was cancelled by Allen and Previn, she proactively asked Lesley Groff for an alternative meeting. When the alternative was set for 4 December 2015, she accepted that as well. Three acceptances in the same 48-hour window.

That the family's contact with Epstein was a single platonic encounter. On the documentary record: the Joi Ito Krebs cycle email of 23 August 2015 indicates that Epstein had been familiar with Oxman's work for at least two months before the supposed first meeting. The August and November 2017 thread filed under EFTA009588 records seven messages from Oxman herself to Epstein over four months, including a $180,000 funding solicitation, a personal gift offer, an invitation for Epstein to attend her 2020 SFMOMA/MoMA retrospective, and Oxman's direct attribution of the existence of her Vespers installation to his support. The 4 January 2019 Brockman document places her, by speaker billing, in the same institutional event Epstein was invited to four months before his arrest.

A dedicated piece, Ackman's elaboration, works through the 2024 thread line by line. The relevant point in the present article is narrower. Ackman's 2024 framing of his wife's contact with Epstein is, on its specific factual claims, contradicted by the documents released in January 2026.

This may be unintentional — Ackman may have been working from the 2019 statement without independent knowledge of the underlying correspondence. He may have understood the 2024 framing in some narrower sense than its plain reading. The question the present article asks is not why he said what he said. It is what is owed to the public when public statements about the conduct of public figures are subsequently contradicted by primary-source disclosures.

What is not in dispute

The documents discussed here are not contested. They were released by the Department of Justice in January 2026 under congressional authority. Their authenticity has not been challenged by Oxman, by Ackman, by MIT, by Pershing Square, or by any party named in them. They are accessible at justice.gov/epstein.

The events they describe — the planning of meetings, the cancellation of meetings, the offer of an apartment, the fourth invitation, the Brockman event — are matters of documentary fact. What they mean is a matter of judgment.

What is also not in dispute: Neri Oxman has not been charged with any crime, has not been credibly accused by any party of complicity in Epstein's offences, and has not, on the documents released, been alleged to have known about specific criminal conduct in progress. Her professional record before, during, and since her contact with Epstein is the subject of widespread admiration and is documented elsewhere in detail. Nothing in this article is a claim about her character, her work, or her culpability under the criminal law. The claim is narrower: that the 2019 statement and the 2024 elaboration do not match the documentary record now available, and that this gap is not a small one.

Closing

A public statement of regret for a misjudgment is one of the more difficult forms of public communication. It depends on what is known at the time it is made, on the speaker's access to the relevant facts, and on the speaker's willingness to be candid about what is recoverable. The 2019 statement may have been the best Neri Oxman could do in a moment of genuine institutional crisis at MIT, working from imperfect recall and against deadline pressure. It is, on its face, sincere in tone. The current authors of this site take no view on her motive in writing it.

The documents released in January 2026 nonetheless raise a question that the 2019 statement and the 2024 elaboration do not answer. The contact with Epstein was longer, denser, and more proactively sought on Oxman's side than her public account has suggested. The institutional setting was Harvard as well as MIT. The funding was 2017 as well as 2015. The most recent documented intersection was January 2019, not October 2015.

The half-life of a public statement, when the documents arrive, depends on whether the speaker chooses to update it. Six years on from the 2019 statement and two years on from the 2024 elaboration, no such update has been issued. The documents are public. The contradiction is on the record. What follows next is, properly, a matter for the named parties to address.

This site exists to make the documentary record accessible, to lay out the contradictions in plain terms, and to give every named party a public route to respond. The contact details for a right of reply are at the foot of every page. We will update this article in light of any response received.

All quoted phrases in this article are taken verbatim from cited primary-source documents and are kept within the editorial copyright limits set out on the Methodology page. The full document text, scan, and DOJ canonical URL for every cited EFTA ID are available on the Sources index. A summary of the editorial standards is available there.

Publication date: 29 April 2026. Last reviewed: 29 April 2026. Editorial responsibility: oxmanrecord.org editorial team.