How a Paid Expert Reversed His View of a Notoriously Flawed Prosecution in the Rape of a Bestselling Author

The upstate New York city of Syracuse seems at odds with itself when it comes to a notorious miscarriage of justice. Nearly five years ago, the district attorney of Onondaga County, William Fitzpatrick, stood up in court and excoriated his county’s decision decades earlier to prosecute Anthony Broadwater for the rape of author Alice Sebold. With the DA’s support, the conviction was thrown out. Today, the same county government and that of its main city, Syracuse, continue to fight a lawsuit filed by Broadwater that seeks financial damages for the years he lost behind bars. The conflicts, it seems, aren’t simply between criminal authorities, who view Broadwater as a wronged man, and civil authorities, who defend the original prosecution. A key expert for the city and county seems to be experiencing an internal conflict of his own — or, at minimum, a dramatic change in opinion. Syracuse’s paid expert, a veteran Pace University law professor named Bennett Gershman, filed a report in the civil suit in December 2025 asserting that the city’s prosecutors “did not engage in misconduct” in the Broadwater case. But a little over a year before that, Gershman told me that prosecutors had “manufactured a case” against Broadwater, calling it “the most heinous kind of prosecutorial misconduct — when the prosecutor is creating guilt.” He went on to say, “‘Misconduct’ is kind of glib in this case. … It’s so much worse than plain misconduct. This is tyranny.” In an interview for this article, Gershman said he changed his mind after delving deeper into the case. “The facts,” he said, are more “complex” and “nuanced” than how he initially understood them. Lawyers on both sides of the Broadwater litigation declined to comment for this article. Certainly, lawyers retain paid experts of every stripe for all sorts of actions. But it’s rare to see an expert take a position in court after expressing a different one to a reporter. “It’s not unethical to change your mind,” said Stephen Gillers, an emeritus professor and ethics expert at New York University School of Law. But, he added, Gershman’s reversal is “an embarrassment and it’s going to undermine his credibility going forward.” A potential jury in the case might wonder what he truly believes. Rebecca Roiphe, a professor at New York Law School, who specializes in criminal law and ethics, offered a similar view. She called it “odd” that Gershman would “be willing to give such a strongly worded comment and then take a position as an expert on behalf of one of the parties. That in itself is problematic. It raises concerns.” She said she views the role of being a commentator for a news story as different from being an expert in a legal case. Commentators should approach the task from a starting point of neutrality, she said. Being an expert, by contrast, has an inherently partisan aspect. “I think it gets confused if you do both,” Roiphe said. ProPublica recently published an in-depth narrative investigation of the original criminal case that examined multiple lapses in the prosecution of Broadwater and uncovered a broader failure in the criminal justice system in Syracuse at the time, which allowed one or more serial rapists to continue their assaults — many of which bore similarities to the one that Broadwater had been convicted of — for years. The original case dates back to the early hours of May 8, 1981, when Sebold, then a Syracuse University freshman, was brutally raped in a park near campus. Initially, the police did not believe her, even though a medical examination and physical evidence supported her account. Five months later, Sebold spotted Broadwater on a busy street and believed him to be her rapist. She reported the sighting to police, and Broadwater was arrested. From the beginning, the case hinged on Sebold’s testimony. But at a lineup, she identified a man other than Broadwater as her rapist. What happened right after that misindentification is at the heart of the current litigation. In the view of the current DA, Fitzpatrick, the prosecution should have halted the moment Sebold picked somebody else: “You know, she didn’t pick out the wrong guy. She picked out the guy,” Fitzpatrick told me for the earlier article. “She picked out the guy that she thought had raped her. And it wasn’t Anthony. Case is over. Stop.” But the prosecution continued. Sebold identified him as her rapist at trial. Broadwater was convicted and ultimately served 16 years in state prison, and lived as a registered sex offender for nearly 23 more. How Sebold described what happened after the failed lineup identification has remained broadly consistent over the years. But there have been different shadings in the account presented in her 1999 memoir about the case and in her 2025 deposition testimony in the civil suit. Her memoir suggests she was influenced by police officers and a prosecutor. In “Lucky,” she wrote that after the lineup she “searched the eyes of the uniformed man for whether I had chosen the right one.” After that, she “felt a wave of nausea” and became convinced she had “chosen the wrong man.” In her June 2025 deposition, Sebold testified that she knew before she spoke to officers or the prosecutor, Gail Uebelhoer, that she had gotten the lineup selection wrong. But she also testified that “there was no way for me to be sure at that time, and then certain things happened that kept reinforcing” that she had picked the wrong man, she said, including a look of disappointment from a detective and Uebelhoer’s remarks to her. These distinctions matter because if police or prosecutors influenced Sebold, it could constitute misconduct. And what happened in those moments is particularly relevant because the prosecution made no attempt to pause the case or investigate further after the failed identification. Uebelhoer had Sebold write an affidavit in which she explained that she picked the man who had been standing next to Broadwater because he was looking at her. They looked “almost identical,” she stated in the affidavit. Uebelhoer then told her, according to “Lucky,” that she had been duped by Broadwater, who had requested that another prisoner be included in the lineup because all the others differed from him noticeably in height or weight. “He uses that friend or that friend uses him, in every lineup they do,” Uebelhoer said. (Both men maintain they had never been in a lineup before. Uebelhoer declined to be interviewed by ProPublica. In a 2025 deposition, she testified that she had little memory of the Broadwater case.) Sebold’s memoir later became a bestseller, and through a tangled series of events that began when producers decided to make a film version of the memoir, the book ultimately helped lead to Broadwater’s exoneration in 2021. After his conviction was vacated, Broadwater sued the state of New York for wrongful imprisonment. The state agreed to pay $5.5 million in March 2023 to settle the case. The city of Syracuse and its surrounding county, by contrast, have so far resisted Broadwater’s claims in a separate lawsuit alleging that they violated his constitutional rights through a malicious prosecution. Broadwater’s attorneys contend that the detective and prosecutor engaged in misconduct by making “false and highly suggestive statements to [Sebold] that led her to identify Mr. Broadwater in court,” and then kept those statements to themselves, which further undermined his defense. That’s where Gershman comes in. As author of a textbook called “Prosecutorial Misconduct,” he is one of the nation’s foremost experts on the subject. The textbook catalogs the ways prosecutors can abuse their powers. He has also warned prosecutors to be wary of eyewitness identifications, citing them as “the largest single source of wrongful convictions.” I had previously interviewed Gershman for a series I wrote on prosecutors who suffered no consequences when they withheld evidence or committed other transgressions. It seemed natural that he’d have insights on the Broadwater case. When I spoke to Gershman in August 2024, I sent him the transcript of the original trial and the motions to vacate Broadwater’s conviction and asked if he could help me identify whether there were any elements of prosecutorial misconduct. After he reviewed the materials (and also read a lengthy New Yorker story about the case), Gershman seemed beside himself. He told me that he had never seen anything quite like it in his 60-year legal career. “I can’t think of a case where a prosecutor has so clearly manipulated the witness into testifying against the person accused of a crime,” he said. “I haven’t seen anything so blatant; so grotesque as what I see here.” That was 2024. Then came his assignment for the city and county and his 2025 report. (Gershman said he notified them at the outset that he had spoken to me.) In his 2025 report, Gershman wrote that Uebelhoer had merely “expressed her opinions” about the lineup and was under no obligation to disclose what she said to the defense. She “behaved properly and professionally, and there is nothing in the record that could remotely be used to undermine her integrity and professionalism.” When I called Gershman recently to ask about his reversal, he insisted that he knew “absolutely nothing about the case” when we first talked and had no recollection of reading the transcript. He noted that he had not yet read Sebold’s memoir at the time of our conversation. His thinking, he said, had evolved as he studied the case more closely. Most important, he said, Sebold hadn’t yet testified in a deposition for Broadwater’s suit. “I don’t think it’s fair to say that I may have made contradictions between what we talked about way back then and what I later learned,” he said. His new opinion fixates on the portion of Sebold’s 2025 testimony where she said she recognized her erroneous lineup pick on her own. In our most recent conversation, Gershman dismissed the account Sebold gave in her memoir and downplayed the parts of her testimony that were more ambiguous. In Gershman’s view today, Uebelhoer’s remarks had no impact on Sebold or the verdict. Anything that the prosecutor or officers said after the lineup was “totally, almost, gratuitous. It didn’t have any bearing on her identification,” Gershman said. He noted that Sebold was asked at trial about her botched identification. A few hours after our interview last week, Gershman called me again, unprompted. He offered what seemed like another zigzag. This time, he told me that Uebelhoer did, in fact, commit misconduct, but that it hadn’t affected the outcome. When I pointed out that his report explicitly stated that the “prosecution engaged in no misconduct,” he said he now wanted to qualify that: “The prosecutors did not engage in misconduct, as I see it, which prejudiced the defendant’s constitutional rights. That’s what I intended to say.” As he summarized it, “She shouldn’t have said what she said, but it didn’t matter.” (Deeper in his report, he also referred to the statements from the detective and prosecutor as “irrelevant and incompetent.”) Gershman emphasized that he had been asked to assess legality, not ethics. His assignment, he said, was to ascertain whether Uebelhoer should’ve disclosed her remarks to Broadwater’s lawyers before trial, not to render a judgment on whether it was appropriate to make them. “I took a legal position that they didn’t have to be disclosed because they didn’t constitute Brady evidence,” he said, referring to the landmark Supreme Court ruling Brady v. Maryland, which requires prosecutors to disclose favorable evidence to the accused. “I don’t do this for the money,” Gershman testified in his deposition, explaining that he was paid $10,000 for the assignment. “I do this because I’m interested in this kind of work. I’m an educator.” Should Broadwater’s civil suit ever reach trial, Gershman will likely be questioned about his evolving positions. If that happens, one challenge will be to convince a jury that his current view is more believable than his previous one.