On the eve of its sold-out annual conference atop San Francisco's Nob Hill, the king of the crypto hill appears to be at its zenith--and closer than ever to being toppled.
Everyone who's anyone in Internet security gathers this week at RSA Data Security's annual Data Security Conference. In fact, the RSA Data Security Conference is so much the place to be seen that 30-plus companies are making announcements today just before it opens. Then, some 3,000 cryptographers, who scramble information with complex mathematical formulas to keep it private, spend the week at seminars, keynotes, bars, and at the popular "Cryptographers' Round Table."
RSA's Web page dubs the confab "the sine-qua-non event of the crypto community," but the company faces challenges to its claim to the title "King of Crypto."
There is no single competitor, no united front of rivals, no real chance that RSA will be toppled soon. Rather, it's because of the sum of several smaller factors that the company faces real threats to its hegemony for the first time:
RSA's patents on its encryption algorithms expire in September 2000.
Elliptic curve cryptography, an encryption alternative to RSA's algorithm that some regard as more efficient and more suitable for smart cards or handheld appliances, is making inroads. CertiCom is licensing its elliptic curve to manufacturers of smaller devices, including Motorola and VeriFone.
Diffie-Hellman, another encryption algorithm, is now available for free because its patents expired in September.
Pretty Good Privacy, privacy pioneer Phil Zimmermann's company and a long-time RSA nemesis, has been purchased by acquisitive Network Associates, itself the result of the merger of McAfee and Network General. Network Associates today outlined its security strategy, and, despite its claims, that strategy puts the Secure One alliance--McAfee, RSA, RSA parent Security Dynamics, and certificate authority VeriSign--up in the air.
In Europe, recent disclosures from the spy community question whether RSA really invented public key cryptography. Recently published papers suggest British intelligence and perhaps U.S. spies created crypto concepts that RSA later patented.
Efforts to get the U.S. government to relax restrictions on exporting strong encryption have stalled. Limits on U.S. companies selling products overseas with strong encryption could hurt RSA and the American software vendors to which it sells. Foreign companies can sell strong crypto anywhere.
Although RSA's secure email protocol, S/MIME, is back on the standards track at the Internet Engineering Task Force, RSA crypto will no longer be required. PGP's secure email protocol, OpenPGP, is farther along at IETF than S/MIME, though both will likely win approval.
Individually, none of these factors puts RSA in jeopardy, as company president Jim Bidzos is quick to point out. They're old news, he argues, insisting his company isn't really influential and therefore he isn't worried about falling from the heights of success.
"We were smart early on, not powerful or influential. I don't know of any way we have any influence," Bidzos said in his "aw-shucks" style.
Most of RSA's revenue comes from crypto toolkits, not licenses for its patented algorithms, he argues. S/MIME won't drop RSA, just add other crypto--and it's already a de facto standard.
The company is certainly flexible. After bad-mouthing elliptic curve crypto for a long time, RSA will add EC to its BSafe toolkit by mid-year, even though RSA algorithms are more battle-tested than the new technology.
PGP isn't a true competitor to RSA, despite years of bad blood (and at least one still-pending lawsuit) between the companies. PGP mainly encrypts email; RSA does crypto engines and toolkits.
The question of whether RSA or some spies invented public key crypto has been around for a decade, Bidzos notes. In my view, that dispute's about bragging rights, not business.
Let's acknowledge RSA's strengths. Through a combination of Bidzos's charm, adroit marketing, and solid crypto, RSA has a strong brand and dominant market share. Now part of publicly traded Security Dynamics, it has access to capital it couldn't touch as a privately held company.
Still, two broad trends in Internet security could buffet RSA: consolidation as bigger players move in, and the growth of smaller devices with limited memory and processing power.
RSA's 1996 acquisition by Security Dynamics was a harbinger of the consolidation theme, but now Network Associates is network security's biggest consolidator. Others will react. The trend is toward one-stop security shopping, where a single firm offers a range of services--including network management, not just security. In that environment, Security Dynamics/RSA may not be big enough, making the company a candidate to buy others or sell out. Bidzos thinks RSA can ride out consolidation on its brand.
Vendors' clamoring about smart cards, handheld PCs, smart phones, home ATMs, and the like creates an opening for other crypto vendors. RSA's algorithms require a lot of processing power, which smaller, cheaper devices don't have. That's precisely where elliptic curve crypto is nibbling away at RSA's mindshare.
Still, don't underestimate Bidzos. RSA may well weather this mini-storm. It may all look like a crypto love-fest in San Francisco this week, but RSA rivals know that if they're going to overthrow the king, it's now or never.
Senior writer Tim Clark conducts his e-commerce love-fest in this space on Mondays.