s chips get cheaper, products get smarter. Sometimes they can get too smart for their own good.

Consider the following examples of products that are capable of limiting how they are used:

¶Some inkjet printers have chips in their ink cartridges that prevent operation if the cartridge has been refilled.

¶Cellphones sometimes come with chips in their batteries that prevent operation if the battery isn't the right brand.

¶Compact discs can have copy-protection systems that keep them from being played in personal computer drives.

Until recently, the after-purchase use of a product has been crudely controlled via contracts, licensing or mechanical design, but now it can easily be controlled through chips and cryptography.

Microsoft recently announced Palladium, a plan for creating secure computing platforms. The Palladium architecture creates an operating environment that allows only digitally signed software to be executed. That should, in principle, help eliminate computer viruses and other sorts of security problems.

But Palladium can also be used for digital-rights management on your PC. This means that only certified programs could be run, and only certified content could be displayed. At the level of bits, censorship and digital-rights management are technologically identical.

Ross Anderson, a computer security expert at the University of Cambridge, described some of the implications of Palladium at a recent conference in Toulouse, France.

What are the economic implications of technologies that can control after-purchase use? The answer depends on how competitive the markets are. Take the inkjet printer market. If cartridges have a high profit margin but the market for printers is competitive, competition will push down the price of printers to compensate for the high-priced cartridges. Restricting after-purchase use makes the monopoly in cartridges stronger (since it inhibits refills), but that just makes sellers compete more intensely to sell printers, leading to lower prices in that market. This is just the old story of "give away the razor and sell the blades."

But if the industry supplying the products isn't very competitive, then controlling after-purchase behavior can be used to extend a monopoly from one market to another. The markets for software operating systems and for music and video content are highly concentrated, so partnerships between these two industries should be viewed with suspicion. Such partnerships could easily be used to benefit incumbents and to restrict potential entrants, a point made by Mr. Anderson.

But there is another set of problems associated with controlling after-purchase use: these technologies can reduce innovation.

Eric von Hippel, a professor at the Sloan School of Management at the Massachusetts Institute of Technology, has documented the importance of "user innovation" in industries as diverse as integrated circuits and mountain biking.

Professor von Hippel's surveys show that roughly a quarter of the users of computer-aided-design software report developing innovations for their own use. One-fifth of mountain bike users do the same thing.

Manufacturers invest heavily in research and development to discover new uses for their products. But the users are often better innovators. After all, the users are closer to the problem: they rely on the products every day for a variety of tasks in a variety of environments, so it is not surprising that they come up with uses the manufacturer never thought of.

Professor von Hippel argues that user innovation is such a strong force that companies should provide tool kits that allow users to experiment with their products.

Such innovation may be sharply curtailed if manufacturers are able to control after-purchase use. Consider the three examples previously mentioned.

A hot area of computer-chip research design involves taking off-the-shelf inkjet printers, loading the cartridges with magnetic ink and squirting integrated circuits onto metalized plastic. That technology may revolutionize integrated circuit production but it definitely requires using products in ways the manufacturer didn't intend.

What about cellphone batteries? There are now hand pumps that allow you to produce enough juice to charge your own batteries. Inventors are experimenting with putting such pumps in your shoes so you can charge your cellphone by merely walking around. This would be great for users, but it is hard to experiment with such technologies if you can use only certain power sources in your cellphone.

Digital-rights management can also reduce innovation. The No. 1 song in England today is a remix of a 30-year-old Elvis B-side single, "A Little Less Conversation." Nike commissioned a Dutch disc jockey, JXL, to do the remix for its World Cup ad campaign. He tweaked the instrumental balance and added a techno back beat to create a fresh new sound.

That sort of thing will be simply impossible if digital rights management becomes commonplace.

One might argue that impediments to user innovation could be overcome by negotiation. After all, Nike got permission from Elvis's estate to do the remix. Surely companies will see that it is in their interest to encourage customers' innovations?

Maybe not. Typically, innovators need to experiment before they approach manufacturers or rights owners. But once they have made the investment to figure out whether the innovation is promising, their bargaining power is reduced. At that point, the rights' owners hold all the cards, and they can choose licensing fees to extract most of the benefits of the innovation. That, of course, reduces the incentives for user innovation in the first place.

Could the innovators approach rights owners before they experiment? Well, anyone can claim to be an innovator. How would the rights owners know who should get permission to circumvent their technology?

Either way, innovation is discouraged, hurting not only consumers but also producers. Too much control can be a bad thing, particularly when innovation is a critical source of competitive advantage