Chow had his own lawyers, a pair of Houston locals by the names of George Murphy and Terry Yates. In the Texas criminal defense scene, Murphy and Yates were local legends, revered as much for their charisma as their unorthodox legal approach. Yates in particular was great with juries, a lifelong Houstonian with a thousand-lumen smile and a pleasant Texas drawl. Immediately, Yates and Murphy focused on changing the case’s jurisdiction from Virginia. They weren’t impressed by the FBI’s assertion that a single microsecond electronic impulse was enough to establish jurisdiction. A federal judge agreed, and the case was moved to Houston. Although the two did not know how to program computers, and this was their first intellectual property case, they were now playing on home turf.

During jury selection, the prosecution struck down any jurors who had ever downloaded music online, even legally. Rivera, Murphy, and Yates did the same, and everyone under the age of 40 was recused from the jury pool. The most important music piracy case in the history of American criminal justice would be decided by a rarefied group of middle-aged Texans still stranded in the compact disc era.

The trial began on March 15, 2010, at a federal courthouse in Houston. When Glover was called to testify he performed poorly. Although he believed Kali and Cassim to be the same person, when cross-examined by Rivera he was forced to admit that he had no concrete evidence to show this was true. If Cassim were to talk in court, Glover could maybe identify his voice. But Cassim didn’t have to talk; the Fifth Amendment guaranteed his protection against self-incrimination, and Rivera could invoke it on his behalf. Amazingly, the FBI did not present recordings of Cassim’s voice as evidence, and, in five days of trial, Cassim never said a word.

Saunders went next. He performed poorly as well—he knew even less than Glover. He couldn’t tie Kali to Cassim, and they had never even talked on the phone. During testimony, Rivera presented the “unsecured wireless” argument, which Saunders found preposterous. The two got into a heated geek-off about how IP addresses were assigned, until the judge told them to cool it.

After this inauspicious beginning, Glover was unsettled. He was still sure that Kali and Cassim were the same person, but now he was beginning to wonder if a jury could be convinced. Could Cassim actually beat this thing? If so, could Glover have beaten the case too? He began to have second thoughts about his own irrevocable decision to plead guilty, and his decision to testify.

The trial continued for four more days. DOJ trial attorney Tyler Newby, one of Jay Prabhu’s deputies, called a dozen more witnesses, including Peter Vu. He entered many pages of server logs into evidence. He presented a paper trail of leaked compact discs. He presented the phone records showing that Glover had called Cassim’s cell phone hundreds of times, the same cell phone Cassim had had on him when he was arrested. Rivera repeatedly disputed this evidence, but for some things—like why Adil Cassim had for several years felt the need to carry on a long-distance relationship with a CD packaging plant employee—he didn’t seem to have good answers.

The trial concluded on March 19. After five hours of deliberation, the jury returned a verdict. Cassim was not guilty. Out of the hundreds of cases brought in Operation Fastlink, it was the first nonconviction. Glover couldn’t believe it. He was going to jail and Cassim was going free, based on an argument over wireless routers. Glover became angry—not with Cassim, but with himself. He never should have signed the deal. He never should have talked to the Feds. He should have gotten a better lawyer. He should have taken the risk. The “unsecured wireless” defense had worked.

Or had it? After all, it wasn’t just Cassim on trial. Matthew Chow had been found innocent as well, even after admitting to being a member of the group. And if Chow wasn’t guilty, the jury must have had something besides router security on their minds. Even as Rivera and Saunders were arguing the finer points of IP address assignment, Yates and Murphy had tried a different approach. Not wanting to bore the jury with technobabble, they instead had focused on the legal definition of the word “conspiracy.” Typically, a conspiracy benefited the conspirators in some obvious way, but in Chow’s case that didn’t seem to have been true. There was no evidence to show that Chow had ever made any money off his participation in RNS. In fact, it seemed he was losing money, spending a portion of his paycheck buying CDs in exchange for—what, exactly? Pirated movies that were in most cases already freely available elsewhere? Yates explained to the jury that Chow wasn’t engaged in a conspiracy. He was just hanging out online with friends.

This argument proved effective. In conversations after the trial, several jurors had said that, while they understood the defendants were probably guilty, they didn’t agree with the severity of the potential punishment, so they had instead decided on acquittals. The legal term for this was “nullification.” It referred to an unusual feature of the American legal system, one that prosecutors and judges tried to keep quiet. Nullification was the prerogative of juries, while accepting a preponderance of evidence, to override laws they saw as unjust. This was the real reason for Chow’s not guilty verdict, and probably Cassim’s too.

For more than a decade Rabid Neurosis had burrowed its way into the music industry’s supply chain. They had scoured eBay for early CDs; they had bribed radio DJs and record store employees; they had sourced moles inside warehouses, and television stations, and music studios; they had even made their way into the factories themselves. They had leaked 3,000 albums a year across every genre. Across the globe they had built a network of infiltration and dissemination. In the shadows of the Internet they had stashed their secret troves of pirated material and kept them locked under uncrackable encryption. A team of expert FBI agents and a small army of private detectives had tried, and failed, to work their way into the group for more than five years. The economic damage they had caused to the recording industry was measurable and real, and ran to millions and millions of dollars.

But on March 19, 2010, a Texas jury, specifically selected for its technological unsophistication, found that the laws that prohibited these activities did not have to be obeyed.

EPILOGUE

Six months after the trial of Adil Cassim concluded, I met Karlheinz Brandenburg in person for the first time. By this time he was 53 years old and his graying beard gave him a distinguished, wizardly affectation. He had continued his work as the director of Fraunhofer’s Institute for Digital Media Technology in Ilmenau, and was an avuncular presence there. His students spoke of him with affection, but he had no children—the mp3 was his legacy.

Now, with the rise of Internet streaming services like Spotify, its retirement was looming. Dieter Seitzer, Brandenburg’s thesis adviser, had anticipated this switch over 30 years before, and his original vision for the mp3 was for streaming media, not storing it. But Spotify didn’t use the mp3 format. It used Ogg, the open-source alternative. Brandenburg and Grill had long suspected Ogg of infringing on their patents, but those patents were more than 20 years old and beginning to expire. The technology was free now. It was the music that cost money.

Still, he could hardly complain, as he was wealthy and successful by any standard. The fruits of his insights could be found anywhere electronics were sold, tracing all the way back to the ancient, five-song MPMan player that Saehan International had commissioned all those years before. In fact, Brandenburg still owned one of those. The thing didn’t work anymore, of course—the battery had crapped out, and no modern computer supported the 20-pin connector by which it transferred files. But for some reason Brandenburg, who was “not sentimental” about technology, had held on to it.