(FORTUNE Magazine) – Near the center of the walled medieval district of Estonia's capital, Tallinn, sits the NoKu bar. It's almost impossible to find, on a cobblestone street behind a pair of old, unmarked wooden doors that unlock only with a magnetic keycard, and up a set of rickety stairs. In Estonian, "NoKu" is an acronym for "young culture"; the private club is full of twentysomethings in jeans, drinking local Saku Original beer to rock music. The bar's name has another meaning: Read as one word, it's slang for "penis." Both the hidden nature and the cheeky attitude of the place fit perfectly with the company I'm here to meet.

Almost a dozen computer programmers and engineers are gathered around a large wooden table in the back of the bar on this bitterly cold mid-December night. They work for a startup called Skype, which produces software that allows people to make free, incredibly clear voice calls from their PC to any other PC in the world. "We're building the next great communications platform," declares Andreas Sjolund, a product manager who has shuttled to Estonia from his native Sweden. Next to him is white-bearded Kaido Karuer, the eldest of the group at 34. The rest include a few Estonians straight out of college and a dark-haired Russian who, when Estonia declared independence in 1991, chose not to ally himself with either Russia or Estonia and is now stateless. Just about all the Skypers are in their mid-20s and perfectly fluent in English. And every one of them is confident that what they're doing will make telephone companies irrelevant.

"We've designed this to be able to support an infinite amount of users," says Sjolund. Karuer tones him down: "We can't do much more than six billion, actually." When asked to name the company's biggest competitor, three at the table quickly reply, "No one."

Normally, bragging Baltic programmers don't inspire much fear among the FORTUNE 500. But these guys do. The reason has less to do with technology or Estonian know-how than with the track record of their bosses. In 2000, Niklas Zennstrom and Janus Friis invented a program called Kazaa. You know Kazaa--the infamous file-sharing program that serves as the global free market for music, video, and porn. Last year "Kazaa" was the top search term on Yahoo, beating out Harry Potter, 50 Cent, even Paris Hilton. With more than 315 million copies residing in PCs from Shanghai to Cheyenne, Kazaa has become the most downloaded program in the world, and it has made Zennstrom and Friis two of the chief enemies of the music business. "They have resulted in significant damage to the record industry," says Matt Oppenheim, head of legal affairs for the Recording Industry Association of America.

Now these two are trying to work their magic again--or deploy their pirate ships, depending on whom you're talking to. This time they're aiming at the telecom industry, and it could get ugly: Since August, more than six million copies of Skype--a product still in its beta, or test, version--have been downloaded, and 2.4 million people have become registered users. On average, 11 new Skypers sign up every minute. And big-name venture capitalists are angling to get in on the ground floor.

As their workers put down pints of Saku, Zennstrom and Friis huddle in their chief programmer's mud-caked Toyota Land Cruiser a few blocks away, plotting the next moves for the company. It's a makeshift meeting--the founders are constantly traveling, so face-to-face talks happen wherever they can--but the decisions they are ironing out will be closely monitored by people in high places.

Very high places. Just 24 hours later and 5,700 miles away, Michael Powell, chairman of the Federal Communications Commission, sits before academics, students, and telecom executives at the University of California at San Diego and explains that telecom as we know it is through. "I knew it was over when I downloaded Skype," he says. "When the inventors of Kazaa are distributing for free a little program that you can use to talk to anybody else, and the quality is fantastic, and it's free--it's over. The world will change now inevitably."

These weren't the guys who were supposed to change the world. The poster boy of music piracy remains Shawn Fanning, the founder of Napster. But Napster is long gone (the new Napster 2.0 shares only the name), Kazaa is still attracting millions, and Zennstrom and Friis are shrouded in obscurity. For now they would like to keep it that way. Skype doesn't list a phone number, and its office location, even what country it's in, is deeply buried. When a public relations representative for Skype reached me, after being forwarded a message I wrote to Skype's general e-mail address asking to meet the men, she agreed on the condition that I not reveal the locations of their offices--information music industry lawyers would love to get their hands on. Two weeks later I was bound for Estonia, Sweden, and London. Where, exactly, I can't say.

My first meeting with Zennstrom and Friis was in a sushi restaurant in Tallinn. The Estonian waitress and the sushi chef wore kimonos. The Kazaa founders don't often speak with the press, and I wasn't sure what to expect--so I expected revolutionaries. What I found were quiet, easygoing, playful businessmen.

They walked into the restaurant together, both late, something investors and friends have grown accustomed to. Both are about 6-foot-4 and lanky, and walk with stooped shoulders. A decade apart in age--Zennstrom is 37 and Friis 27--the two switch seamlessly between acting like business partners and like brothers. Zennstrom is the leader, doing most of the talking, especially when it comes to business and strategy. When we meet, he sits down, kicks his long legs out in front of him, and starts running through numbers: the size of the telephony market, the number of broadband households in the U.S. and Europe--and how his startup can offer free phone calls forever and still make bundles of money by charging for things like voicemail or a connection to a regular phone, or through reselling arrangements with ISPs and hardware makers. While Skype draws on some of Kazaa's technology, this isn't file sharing, it's just conversation via the Internet, so it's hard to imagine anything to sue about. "There is multibillion dollars in potential for Skype," Zennstrom says. "We're not here to try to make some small business."

If Zennstrom is the left brain, Friis is the right. When he chimes in it's usually to rib Zennstrom or emphasize one of his points. Zennstrom once spent an hour detailing Skype's business plan to two U.S. venture capitalists, summing up: "In a nutshell, that's what we're doing." Friis, who had remained silent almost the entire meeting, said, "A very, very long nutshell." His customary uniform--jeans, sweater, backpack--contrasts with Zennstrom's suit and rumpled shirt. He's the innovator who doesn't have time for business niceties.

In the dot-com days, executives of major companies trembled at the thought of tinkerers in a garage inventing some business that would destroy them. Then the boom went bust, and the curtain was pulled back to reveal a garage full of sock puppets and Razor scooters. For big companies, today's bogeymen are whistleblowers and scheming competitors. Yet the threat of disruptive technology remains--and Zennstrom and Friis are working hard to prove it.

Still, they don't see themselves as rebels. It's not in their demeanor or their attitude. "I don't think they're grinding an ideological ax," says Ian Clarke, a friend and technologist who developed the file-sharing system Freenet. "Yes, I think what they have done has been disruptive, but I don't think they do things because they're disruptive. They do what they think will make a successful business. When you are doing something disruptive, it typically means there is a business opportunity there."

The two met in 1997 when Zennstrom was working for Tele2, an upstart telecom company in his native Sweden. The son of two teachers, Zennstrom decided early on that he wanted to create something, not teach about what others had done. He earned a dual engineering and business degree at Sweden's Uppsala University, completing his final year at the University of Michigan, before joining the small phone company. As one of his early jobs, he was sent to Denmark to build an ISP business. Aiming to form a small and nimble group, he put out an ad in a few Copenhagen papers. It caught Friis's attention.

There was little to indicate he was cut out for the job. At the time, Friis was doing customer support at a competing ISP, a surprising job for someone as blunt as Friis. He had quit school at 16--"I didn't feel like much more education," he explains--and left his parents' home outside Copenhagen to travel through India. At the Tele2 interview, Friis and Zennstrom had a meeting of the minds about strategy, and Friis was hired as part of a four-man team. Over the next two years, Tele2 kept Zennstrom country hopping, moving him to Luxembourg, then Amsterdam. With each move, he'd import Friis.

By late 1999 both felt as though they were missing the dot-com train, and Friis prodded Zennstrom to quit Tele2 so they could start a company together. Why help a big company get bigger when they could found something themselves? As for what they'd make, well, they'd figure that out later. But the time to do it was now. Zennstrom turned the kitchen of the Amsterdam apartment where he and his wife lived into a makeshift office, and Friis moved into the guest room. "We knew we'd come up with something," says Zennstrom. "Or at least we hoped we would."

They tossed around--and out--dot-com-mania ideas such as a product-review site. Instead they built from experience. While crafting the ISP business and an accompanying portal for Tele2, Zennstrom had grown frustrated about having to buy enough bandwidth between the U.S. and Europe to enable subscribers to watch movie trailers or listen to streaming music. The two wondered if there wasn't some way to make money by storing files locally--even better, why not store them on subscribers' computers instead of their own?

Like most of Zennstrom and Friis's ideas, this one wasn't original. The concept of distributed computing--or peer-to-peer (P2P)--had been around since the early days of the Internet. The technology allows people to access one another's computers or tap someone else's machine for, say, storage or computing capacity. In 1999, P2P leapt from geeky obscurity to how-can-I-invest obsession with the launch of Napster. Swapping songs over Napster became the national collegiate pastime, and by December the RIAA had filed suit.

Zennstrom and Friis were as caught up in P2P as the rest of the wired world but insist that they never thought they were making a replacement for Napster. They say they wanted to make something that would enable people to share photos, text files, jokes, whatever. But what they eventually built was an almost unstoppable version of Fanning's software. (The RIAA's Oppenheim calls their professed early innocence "bull.") They weren't sure how they'd make money; that would come later. "What I learned when working with Tele2 is that sometimes when you come in with a business plan, it's like, 'Why are you wasting your time writing that? Just go out and do it,' " Zennstrom says.

Neither Zennstrom nor Friis was a programmer, but they knew where to turn. While at Tele2, Zennstrom had placed an ad in Tallinn's Eesti Paevaleht newspaper asking for programmers to bid on creating a portal for the company. (The headline for the online ad: "Supermodels not needed! We want your brain!") Applications were due within a week. Four computer-game programmers in Estonia who had formed a company called Bluemoon saw the ad, spent the weekend learning the required computer language, and by the end of the weekend had a complete working design for the portal. Zennstrom hired them immediately. Now, gone from Tele2, he turned to them again.

The Estonians knew exactly what he was looking for. In the 1960s, as part of the U.S.S.R., Estonia had poured funds and people into its Institute for Cybernetics. While Cybernetics centers in other republics focused on biosciences or math, Estonia's devoted itself to computer programming. Ahti Heinla, one of the Bluemoon men, had been taught programming at age 10 by both of his parents, who worked at the institute. After the country declared independence, it quickly put its tech savvy and relatively cheap wages to work. Combine a developed distrust of centralized authority--a few decades of Soviet central planning will do that to you--with a libertarian streak that runs through the country, and it's easy to understand the direction Kazaa took.

The Bluemoon programmers, working closely with Friis, crafted a system that thumbed its nose at authority and worked ingeniously. Napster's servers kept a kind of card catalog of what files sat on which users' computers. That meant that as the number of users increased, Napster had to add servers. Worse, the servers enabled the courts to prove that Napster was involved in the illegal trading of copyrighted material. The Bluemoon system, dubbed FastTrack, avoided all that. Computers that were logged on to FastTrack negotiated among themselves to find the ones that were the fastest and had the best Net connections. Those so-called supernodes automatically took on the task of hosting the master card catalog. As more users signed up, more of these supernodes were created. FastTrack essentially outsourced its server needs to its users. Plus, if music files were traded, the FastTrack system prevented Zennstrom and Friis from knowing exactly what was going on.

On the business side, Zennstrom and Friis decided to license access to FastTrack--akin to building the Internet and charging browser makers a fee to use it. They hired programmers in Romania to write their own "browser," and while putting back Thai beers at an Amsterdam restaurant called Sawaddee Ka, they decided to name their product after the place. And so Kazaa was born.

Their timing was impeccable. The RIAA knocked out Napster in July 2001, when a court effectively ordered the service shuttered. There were other Napster alternatives, but many had been built by geeks for geeks and were difficult to use. The music industry thought it had dodged a bullet. Yet almost immediately, analysts at Big Champagne, a company that tracks music-file sharing, saw traffic spike over the easy-to-use Kazaa and the other FastTrack services developed under license, called Morpheus and Grokster. Zennstrom and Friis thought they could work out licensing deals with entertainment companies, creating a fully legal system in which artists were paid.

In fall 2001 they jetted to the U.S. to try to make it happen. Holed up in a Motel 6 near Los Angeles, Friis logged on to the web and checked a site called Dotcom Scoop. Posted there was an internal memo from the RIAA, a group they were scheduled to meet with the following week to explore licensing deals. The 2,000-word document detailed how Kazaa and FastTrack worked--and ended with a recommendation that the RIAA sue Kazaa's founders for copyright infringement. "The claims are not as strong as those against Napster," explained the memo, "but they are also not so remote as to be wishful."

Zennstrom and Friis immediately called their lawyers in the Netherlands, who told them to abort their meeting--it might be a trap to subpoena them. On Oct. 2 the Motion Picture Association and RIAA member companies--including Bertelsmann, Disney, MGM, and Time Warner (FORTUNE's parent company)--sued Kazaa and both companies licensing FastTrack. Twenty-four hours later Buma/Stemra, the Dutch music-rights licensing group, canceled negotiations with Kazaa about a licensing deal and threatened to sue.

Zennstrom and Friis found out on Oct. 4 just what a hornets' nest they had poked. Instead of canceling their meeting with the entertainment-industry representatives, they sent their lawyers, then hopped into a rental car to cruise around Beverly Hills and await word. About an hour and a half later, Kazaa's lawyers called and told them to come in--the two sides had brokered an agreement guaranteeing that they wouldn't be served for the duration of the meeting. Suddenly Zennstrom and Friis found themselves face to face with a team of music and Hollywood industry lawyers haranguing them for creating a forum for stealing music, and insisting they shut it down.

The meeting ended without any agreement. Oppenheim, the RIAA lawyer who has taken on the role of chief Kazaa-hunter, says the men squandered a huge opportunity that day when he offered them a "pretty big give": Stop the piracy, and he'd drop the suit. They said they wouldn't. "Not that they couldn't," he points out. "But that they wouldn't ... I said, You're going to end up in the same place as Napster."

For Friis and Zennstrom, the whole experience still seems unreal. When Kazaa was launched, both thought they would become hits in Europe--signing deals with European ISPs, working with Buma/Stemra to make a pay-for-share service, building an aboveboard business. Instead they had become adversaries of some of the biggest companies in the world with a brand that was as common in dorm rooms as a bong or a backpack. They felt like the co-stars of a bad comedy. "We look around sometimes and say, 'Where's the camera?' " says Zennstrom.

Kazaa had gone from risk to liability. In early 2002 a Dutch district court ordered Kazaa to shut down--something the Kazaa founders had argued was impossible. It was time to get out.

The men began a series of incredibly complicated legal maneuvers to distance themselves from their creation. They sold the Kazaa program and Kazaa.com site to an entrepreneur in Australia. Her company, Sharman Networks, is incorporated in Vanuatu, a group of islands in the South Pacific best known as an international tax haven. The RIAA insists the sale was more of a handoff. Zennstrom bankrolled the purchase, and Sharman repaid the loan from Kazaa revenues. To deal with FastTrack, Friis and Zennstrom started a company called Joltid, transferred the source code there, then granted Sharman a worldwide license to use it. (Sharman executives declined repeated requests for comment. "They're really not game to talk about" Friis and Zennstrom, says a spokesman.)

Last, as part of the sale, Zennstrom and Friis, along with a Los Angeles software firm called Brilliant Digital Entertainment, created a company called Altnet and declared that it would sell licensed digital files over Kazaa. Today Joltid owns 20% of Brilliant stock. Altnet also pays Joltid $30,000 each month for the license to certain software.

The deals have given Zennstrom and Friis an income stream--albeit a small one for creating the world's most downloaded program--and kept the RIAA and Motion Picture Association's lawyers running in circles. A federal judge threw out most of the RIAA's case against Grokster and Morpheus last summer. The RIAA is now suing Sharman--and Zennstrom and Friis personally. The two have refused to show up in U.S. courts, and Kazaa has been declared "in default."

Not that the RIAA has given up. Late last October, according to a motion filed by Zennstrom, he and his wife went for a morning walk in London's Branham Gardens when they were surrounded by men--including one sporting a black-and-white leather suit and riding a motorbike--who tried to shove a subpoena into Zennstrom's hands. (The RIAA insists there was only one person attempting to serve him.) Zennstrom and his wife took off running.

With the RIAA on the prowl, it's no surprise that Zennstrom and Friis try to keep their whereabouts under wraps. Here's as much as I can say about their Estonia offices: There's not much to say. There are no skateboards lying around, or PlayStations, or dartboards featuring the CEOs of Verizon or SBC. Just a few potted plants and some scattered desks. Sure, the two men are making the most of their rebel status--there is some mention of Kazaa on almost every page of Skype's website--but this is for public consumption only. Inside, they're hoping to build Skype into a business that will make them rich--and along the way make some waves.

Today Zennstrom and Friis are having a rare face-to-face meeting with the rest of the Skype team. In addition to 28-year-old product manager Sjolund, there's the frenetic Geoffrey Prentice, 28, an American who runs Skype's business development, and Toivo Annus, 31, an intense, sarcastic Estonian who chain-eats ramen and oversees programming. Typically this group works together, e-mailing, Skyping, or telephoning Zennstrom or Friis--Zennstrom for money and business-development issues and Friis for product brainstorming. Right now, along with a PR handler from Arizona, they're talking over what the next version of Skype will look like, how it will work, and when it should be launched. Programmers --some working for Skype, some for Joltid, and some unable to say for sure--hunker over their computers.

So far Skype offers PC-to-PC calling. Download the software onto your computer, and up pops a window familiar to anyone who uses instant messaging. You can use a directory to search for other people who have registered with Skype. Click on a user, and his computer starts ringing. As long as both users have microphones, you've got a crystal-clear phone call. A downside: Both parties have to be hooked up to the Net. But trends are on Zennstrom's side. Over the next few years, high-speed wireless Internet, called Wi-Fi, and Wi-Fi chips will spread, making Net access nearly ubiquitous in laptops, PDAs, and cellphones.

The program's real edge comes again from the tricks of the Bluemoon programmers. Like Kazaa, Skype is P2P: The heavy lifting of converting the sounds to packets that can be sent over the Net, routing them, and maintaining the directory of users is handled first by the user's computer, then by other computers on the network. The process seems to work well. While in Stockholm, I called my wife in New York City over Skype using a dial-up connection; the sound was a step above a cellphone's and the cost infinitely better. The P2P underpinnings also mean that Skype can grow without adding much, if any, infrastructure. It costs Vonage--the top provider of paid Internet telephony, with 97,000 users--almost $400 to add a new customer. It costs Skype one-tenth of a cent.

Furthermore, Zennstrom and Friis have the Kazaa name to help spread their message. Not only has Skype been downloaded more than six million times in six months, it is regularly used in more than 170 countries--and it has spread by word of mouth alone. This year Zennstrom wants the downloads to hit 25 million. Skype does have a plan for making money: It will charge for added features like voicemail, connections to regular phones, or placing targeted ads on the main window. But for now, Zennstrom's plan is just to grow like mad first and make money later.

Skype's venture backers like what they see. Bessemer Venture Partners and veteran Silicon Valley VC Bill Draper and his son Tim have all sunk money into the firm. The younger Draper, managing director of VC firm Draper Fisher Jurvetson and a personal investor in Skype, likens the startup to one of his biggest successes: Hotmail. "I like the idea of being able to reach millions and millions through viral technology of any kind. I think Niklas is an extraordinary guy. He has been through so much with his first deal and taken on an extraordinarily powerful industry, and he's going to do it again. This going to be the next great company."

The fame of Kazaa might bring money and subscribers, but it also brings problems. For one, the telecom industry isn't about to let itself be Kazaa'ed. Last winter became the season of voice over Internet protocol, with everyone from AT&T to cable companies like Time Warner announcing that they were offering Internet calling. Public VOIP companies like 8X8 and Net2Phone saw their stocks shoot upward. Voice calling began to look as though it would become just another service over high-speed pipes, no different from video or music.

Plus, all these companies were one step ahead of Skype: Their services worked with regular phones--no PC needed. "What Skype is doing is like a toy," says Hossein Eslambolchi, AT&T's chief technology officer and president of AT&T Labs. "They will realize they can't scale it, they don't have a brand like the AT&T brand, and they don't have the local footprint, which we have. It's going to be very hard to compete with someone like AT&T."

Yet even if Skype can't compete--and the jury's out on that--it can still serve as a massive spoiler. It's getting millions of people used to the idea of getting their voice calling free. That's a hard habit to reverse.

Friends of the men simply marvel that Skype, with minuscule revenue, is even getting any attention from the likes of AT&T. "They're basically kicking the nail of the smallest toe of a giant," says Morten Lund, an early investor in Skype and the CEO of antivirus company Bullguard.com. "And they've done that with, what? One press release."

Air Estonia is going through heavy turbulence. We're on our way from Tallinn to Stockholm, where Friis and Zennstrom keep the second of their three offices--again in an undisclosed location. Zennstrom is tightly gripping the armrests. Friis, sitting next to him, ignores his friend's fear; he's used to it. "Janus thinks I should try skydiving," says Zennstrom.

"Of course you should," says Friis. "Either you'll die of a heart attack, or you'll never be scared to fly again."

The same spirit seems to motivate the venture capitalists who are now knocking on the pair's door (or more likely sending blind e-mails). When the two started Kazaa, they couldn't get a call returned--Zennstrom funded the company with his savings. Now, with Zennstrom and Friis being attacked by the courts and Kazaa out of their hands, investors are eager to be involved in their next wild ride.

Two venture capitalists are waiting in Stockholm. Representing a medium-sized firm, they've flown in from San Francisco for 48 hours, just to meet with Zennstrom and Friis. They spend a few hours talking but leave empty-handed. Skype has already closed an early round of financing, raising something less than $25 million, and is oversubscribed for its next round.

The challenge Zennstrom and Friis must figure out is how to capitalize on their maverick status but still become part of the mainstream. Yahoo Internet Life magazine once dubbed Zennstrom a "Robin Hood of the Net." Yet ask him whom he admires in business, and he'll mention Bill Gates and Steve Jobs, not a free-software missionary like Linus Torvalds, creator of Linux. "This is not a crusade that's driving us," Zennstrom says over drinks. He's out to make a real business.

Which is where London comes in. The two have picked the city as the new headquarters for Skype. It's a global hub, easily accessed by Americans and Europeans, and can't be considered fringe. Right now the office is little more than a rented suite--that much I can say--but the plan is to move into something bigger and to start hiring: MBAs, telecom veterans, salespeople. Zennstrom and Friis want to create a business that lasts--something that Oppenheim and his RIAA crew made sure they couldn't with their first creation.

But then, Zennstrom and Friis aren't bitter. Without Kazaa, they would likely be just two European entrepreneurs trying to get some attention. These days they've got more than they can handle. In conversations, Zennstrom calls the continuing legal problems with Kazaa his "legacy issue"--a techie term that refers to old computer systems. Friis is, if anything, amused: "We were doing all of these partnership deals in Europe, we were based in the Netherlands, yet the entertainment industry still managed to drag us into a huge, lengthy, and costly battle in Los Angeles," he says. Then he laughs: "It's impressive, actually."

Friis can smile about it now. After all, the future of Skype looks much more promising. The threats to this new company are business-related, not legal. Friis doesn't know if he'll stick around long enough to see where Skype goes--his wanderlust leads him to believe that in another three to four years he may exit and leave the company to Zennstrom and professional management. By then, if all goes as planned, Skype will be the kind of business that has an actual address, and the founders will be able to come in from the cold.