Cornell Assault Inquiry Gets State Prosecutor as Evidence Gap Persists

New York has put the criminal investigation under different leadership. On October 1, 2026, Governor Kathy Hochul appointed the Office of Attorney General Letitia James as special prosecutor to investigate and, if warranted, prosecute. Cornell had agreed the previous evening to retain outside counsel for a separate examination of the university’s handling of the case. One concerns potential crimes; the other concerns institutional responsibility. Neither action establishes guilt or misconduct.⁠104⁠109⁠3

We reported on September 30 that Tompkins County District Attorney Matthew Van Houten had reopened the inquiry, that no criminal charges were reported, and that Cornell had not committed to the outside review Hochul requested. That last point has changed. State-level control of the criminal investigation is a further change—not merely another announcement that the local inquiry is reopening.

Doe’s civil lawsuit names Cornell and seven men she accuses of drugging and sexually assaulting her. It alleges pressure to take ketamine, smoke marijuana and drink alcohol until she could not consent, and alleges that fraternity members used a group chat to facilitate further assault. She uses a pseudonym because of concerns about further harm to her life and privacy. These allegations have not been established in court.⁠11⁠10

The reported police records distinguish some initial activity she considered consensual from later activity she describes as unwanted. They describe repeated refusals, increasing intoxication, feeling unable to make decisions and conduct she characterized as coercion. CNN also reported an account of her being physically hit. Those distinctions matter: consent to one act does not establish consent to every subsequent act.⁠15⁠14

CBS News reported obtaining an interview transcript containing Doe’s words: “I can say with 100% confidence, I was raped.” Van Houten told CBS that passage was absent from the sworn six-page summary Cornell police submitted to his office in November 2024. CNN reported the detailed interview dates as November 14–15, 2024. NBC had not independently reviewed the transcript; NPR had not obtained it, though Doe’s attorney, Thomas Giuffra, confirmed her allegation.⁠12⁠14⁠1⁠3

The competing delivery accounts are explicit. NBC News’s October 1 report carried Cornell’s statement that it had turned over a large evidence file. CNN’s October 2 report carried Van Houten’s account that he received a six-page statement and one Snapchat video, not the larger collection. Cornell previously said campus police shared all investigative information with prosecutors. No complete, dated inventory reconciles what was sent, received, identified or examined.⁠1⁠2⁠12

The scale of the university file makes that comparison urgent, but does not prove deliberate concealment. CNN described New York Times reporting that Cornell’s internal investigation assembled more than 1,000 pages. That does not establish that every page existed when police first consulted prosecutors. Campus disciplinary and police investigations are different processes, and later findings cannot automatically be treated as information available during the initial charging review.⁠2

Van Houten’s subsequent distinction is narrower than his initial characterization: particular words were missing from the supplied summary, but he had not read Doe’s full detailed account. The reported interview passage challenges a broader suggestion that she never told investigators she was raped. “I can't explain why additional information did not make it to my office,” he said in the MS NOW account. A document-by-document comparison is needed to determine whether material was not sent, not received, inadequately identified or not examined.⁠3⁠115

Our September 30 report also recorded Van Houten’s correction of his earlier statement that he had not received a Snapchat screenshot. He acknowledged misspeaking while maintaining that possession of it did not change his assessment of consent. That correction remains important history. It is not confirmation that he had the complete group chat, the later-described video, the interview transcript or Cornell’s internal file.

Van Houten defended his original decision on the information he said he received, emphasizing conscious participation, consent to parts of the encounter and voluntary drug use. But consciousness alone does not resolve capacity to consent throughout an encounter. CNN’s legal discussion distinguished voluntary intoxication from substances administered without consent and emphasized assessing particular acts against their evidence. Neither Doe’s allegation nor the prosecutor’s earlier interpretation settles those questions.⁠4⁠14

Other reported records add evidence to examine, not a verdict. CNN described a text attributed to Doe saying 90% of the night was not consensual, a rape-kit request four days after that text, and a pelvic examination two weeks after the incident. The percentage is her characterization, not a forensic measurement. No kit result or complete clinical chronology is established, and the visit descriptions cannot automatically be treated as either one encounter or two distinct encounters.⁠2

ABC News reported three men denying involvement and challenging the complaint as incomplete, updating the two public denials in our earlier report. They invoked Cornell’s Title IX disciplinary process as having cleared them in some respect. That does not establish responses from all seven or identical findings concerning them. Another defendant’s silence is not an admission.⁠9

Defense attorney Jeremy Saland told CNN his client was not accused of sexual activity with Doe, was not found responsible for sexual assault and was neither suspended nor expelled. He described a Snapchat-related finding resulting in an essay or classes, and cited a hair-follicle test as supporting a denial of ketamine use. The complete decision and test result are not independently established. His account nevertheless underscores why the seven men cannot be assigned identical conduct.⁠14

The campus-outcome accounts also remain unreconciled. On September 30, FRANCE 24 English reported two expulsions and five students resuming studies after an essay, workshop or short suspension. CBS News’s September 30 account, citing Cornell’s president, gave two expulsions, two multi-semester suspensions, two findings of no responsibility for sexual misconduct and one student graduating before the inquiry concluded. No verified explanation establishes whether the difference concerns timing, sanction categories or later enrollment. Campus discipline is not a criminal verdict.⁠11⁠12

Hochul’s Executive Order No. 64 identifies the alleged incident as October 19–20, 2024, and makes prosecution conditional on the investigation warranting it. AP reported that she had lost faith in the local prosecutor’s handling. The appointment puts James’s office over the case, but the precise provisions governing the local office’s continuing assistance, record transfers and relinquishment of authority remain unestablished. No completion deadline can be asserted from the available order text.⁠104⁠109

Cornell’s outside-review commitment is real; its implementation remains unverified. Hochul said she spoke with President Michael Kotlikoff before announcing the agreement. No identifiable outside reviewers, charter, budget, publication commitment or binding timetable are established. Independence will depend on who selects and supervises the reviewers, what records they can obtain and whether readers get to see what they find—not simply their location outside the university.⁠1⁠3⁠38

Broader survey figures cannot resolve this case. A secondary summary gives 15% of students reporting nonconsensual sexual contact since entering Cornell in 2025, versus Cornell’s confirmed 11% in 2023: four percentage points, not 4% relative growth. That overall-student measure cannot substitute for an undergraduate rate. Exact subgroup baselines behind reported undergraduate increases remain unverified, and survey participation fell sharply. These findings describe reported campus experience, not proof against any defendant.⁠88⁠102⁠99⁠100

As of the October 2 reporting, no completed prosecutorial interview with Doe and Giuffra, grand-jury presentation, grand-jury outcome or criminal charging decision is established. Van Houten’s earlier plans must now be read alongside the state appointment, not as though he alone controls the next stage.⁠91⁠93

The test of both interventions is straightforward: collect the complete record, establish who examined what, and make decisions the evidence supports. Moving responsibility is consequential. Demonstrating that the full account has finally been considered is the work still ahead.