When is it time to say "no" to innovation? When are innovations so destabilizing that society must intervene? To raise this question isn't to launch down the path toward Luddism. The idea that innovators work under limits isn't new. The great military innovations of World War II spawned legal and practical prohibitions on what could be done with nuclear and biochemical weapons. Medical and pharmaceutical technologies are routinely regulated. So are cars, planes and other transport innovations.

Yet information technologies are often presumed to be in a different category altogether. And there are good reasons to think that's true. Consider two examples. Britain's Labour government has introduced legislation that would make it a crime to withhold a computer password from the country's domestic spying agency. The government also wants its snoops to have the right to read any e-mail and monitor any Web site without even asking for a court order. The U.S. government, meanwhile, limits the quality of cryptographic products-software that encodes secret messages-that can be sold abroad by U.S. companies.