If you own a GPS cell phone that tracks your location, it may already have been used in smart-phone applications that monitor traffic congestion and figure out ways to bypass logjams.

And as long as your personal information wasn't accessible to the company measuring rush hour traffic, you might not mind that much.

But what if someone were to invent an application that would send a message to your phone if you had been within 3 feet of someone who was later diagnosed with the swine flu in an emergency room?

Would you mind getting such a call? If you were the flu patient, would you mind that information being spread around?

Those are the kinds of privacy issues that are mushrooming as more and more cell phone and computer programs use data gathered from individuals to monitor traffic, study social interactions, market products and track health trends.

It's also what prompted Carnegie Mellon University computer science professor Tom Mitchell to write an essay on the issue in today's edition of Science magazine.

The information being swept up today primarily comes from cell phones and computers, Dr. Mitchell said in an interview this week, but also includes security cameras, ATMs, store cash registers and digital TVs.

"As more diverse sensors become pervasive, wireless networking becomes more widespread, and new [computer software] algorithms are developed, a global sensor network monitoring much of humanity might emerge," Dr. Mitchell writes.

While researchers also are working on software designed to protect people's privacy in this burgeoning cyber universe, it is also vital to have "a public discussion about how to rewrite the rules of data collection, ownership and privacy to deal with this sea change in how much of our lives can be observed, and by whom," he writes.

"There are some very positive things that we can do with all this data," Dr. Mitchell said in the interview this week, "and some things I wouldn't want to see done with this data, but there are also privacy issues like we've never seen before."

Dr. Mitchell, a pioneer in designing software that can detect trends and patterns in massive amounts of data, said he didn't write the essay because he knows the answers to these privacy questions, but "to stir up a discussion on this issue."

As a first step, he said, America might want to adopt some of the policies that exist in European nations, which basically say that individuals own their personal information and must give consent ahead of time for its use.

In the United States, he said, it is somewhat more common for companies to start gathering data from individuals, tell them later how they plan to use it and then give them the option of switching to another company if they don't like the policy.

Complicating matters is what happens when people's personal data intersects with the legal system, he said.

As an example, Dr. Mitchell said his iPhone will mark the geographic location of every photo he takes with it, but asks him each time if he wants to skip the "geo-code."

"However," he said, "if the court system decided to subpoena that data, it wouldn't matter what I answered."

The challenge, he said, is that "the laws and the decisions about what could be subpoenaed were developed before we entered this age where data really does persist. If you have a phone conversation these days, there's a record.

"And if the next app I download on my phone is one that records the audio of my 9-to-5 workday, which it could easily do, then all of a sudden all of that would become subpoenaeable."

Because the use of personal information ultimately involves people's civil rights and legal obligations, the federal government may have to step in to provide oversight on these issues, he said.

In the meantime, many scientists are working on programs that allow individual data to be aggregated without divulging personal identities.

One example is health studies that use data from several different hospitals. It is possible to pool that information without revealing personal data, he said, by using cryptography techniques to transmit the data to those who are assembling it.

To explain how such programs work, he gave the example of four people sitting around a table who are curious about how much each of them earns, but who don't want to disclose their salaries to each other.

One way to deal with that conundrum, he said, "is that you make up a very big random number, say 2,453,752, and you add your salary to it, and you pass it to the person on your left. He gets this big number, he adds his salary to it, and he passes it to the next person, who does the same, and he then passes it to me. I add my salary, and when it comes back around to you, you subtract out the random number, and you divide by four, and now we know what our average salary is -- but the interesting thing is, no one knows anything specific about anyone else's salary."

Studies that are looking for average trends can use similar algorithms to get the answers they need, without the researchers knowing the specifics of individual data, he said.

Dr. Mitchell stressed that he wants to see individuals' data be used for society's benefit.

But, he wrote in Science, "until these [privacy] issues are resolved, they are likely to be the limiting factor in realizing the potential of new data to advance our scientific understanding of society and human behavior, and to improve our daily lives."

Washington correspondent Daniel Malloy writes the "Pittsburgh On The Potomac" blog exclusively at PG+, a members-only web site of the Pittsburgh Post-Gazette. Our introduction to PG+ gives you all the details.