Threat to cyberactivism

Southeast Asia will provide endless opportunities for research contrasting and comparing the use of the Internet in different socio-political climates, range of political change, the Internet as an instrument of civil society, and more, using examples from the past, present and future. There are so many variables to explore, not only cultural, economic, religious, censorship, governmental control, citizen apathy or activism, degree of autonomy of civil society, but also the fortunate (or not so fortunate) timing of events.

A bright spot for cyberactivism is that even in Burma where there is such severe control and harsh censorship, civil society organizations are being supported in the global cyberspace by independent press and opposition groups operating their websites from beyond that country’s borders.

Another shining ray is in Indonesia’s history – the Internet conveniently became available just before and during the Asian financial crisis. People could convey information about Suharto’s regime, which was ironic due to the fact that the Internet was brought in precisely as a measure to control the population. This is a good example of the Internet being used for political reform.

The Philippines (2001) also have an interesting story of cyberactivism when President Estrada seemed to escape impeachment by devious means. The Internet became a source of information that organized the resistance. But in their case, they extended the reach of the Internet using cell phones and text messaging, making if difficult for authorities to handle or respond in a timely manner. They could organize quickly and in large groups.

Malaysia in 1998 also used the Internet for “reformasi.” Though it did not have the huge impact Indonesia experienced, it did cause political change.

But what is frightening now, though, is the threat to cyberactivism by data mining and surveillance in the post 9.11 environment. Lawrence Lessig tells us what the difference is regarding surveillance in the computer age. It is the ease in which data is retrieved, stored, and available to be searched at a future time. This will increase in out daily lives as more interactions and transactions take place electronically and become embedded in our media.

These records stay inside the computers belonging to corporations and government agencies. Browsing on the web creates more and more information about you in which marketing service providers are using technology to keep a growing record on us. What could be more Big Brother when our financial transactions and website visits are permanently available, made possible by sophisticated software.

As like the Internet was funded by U.S. federal government research and development, so too is this capability, being built on similar concern for military capability or “strategic intelligence” from bits of data stored in computers around the world.

Oscar Gandy Jr, in his 2002 paper on Data Mining suggests data mining is an applied statistical technique. This extraction is increasingly being automated in ways that are less risk to labor and more risk to society. Global retail chains (he cites Wal-Mart as an example) have already invested in this development to catch the details and extract the value in data being generated daily through their network.

There is pressure on our health care system and government agencies to gather their data in standard form similar to the UPC code so there would be standardization and comparability across transactions.

Most upsetting is data mining efforts are directed towards the “generation of rules” in order to classify objects or people, assigning them to particular classes or categories which would facilitate economic discrimination.

Another form of analysis would be “associative rules” to find patterns of association between demographic characteristics and commercial behavior. “Discriminant analyses” enables contrasting high value with low value customers.

One of the most sophisticated forms of data mining result in neural networks which imitate how the brain processes information. These systems can “learn,” becoming more accurate as they go along because it uses a statistical learning model that applies different variables according to correct or incorrect predictions. One use for this is to support fraud detection, or perhaps data enabling help in choosing one strategy over another, whether in the marketing, political, or even personal arenas.

Gandy points out that while this seems incredible now, 5 years from now this “business intelligence” will have moved “to the average desktop” like Word and Excel have already done. The leaders of this technology digiMine, Accrue, NetGenesis and Personify which sell analytical services to web-based customers. This also includes familiar providers of statistical software like SPSS which include neural networks and rule induction features.

Also interesting to note, the U.S. government is not only providing incentives, but has launched a major effort to speed up this development so it can be “deployed” within 12-18 months – the goal being “ideas to identify and track down “suspected terrorists and predict their future behavior,” and an integrated information base, data mining tools, and analysis aids. (The U.S. military uses satellites for its Internet use.) This expense in the event they will be able to locate a terrorist who buys explosives and visits a website regarding building demolitions. So why do we care?

We will be identified and classified into distinct groups for the purpose of discrimination for the purpose of economic gain. Marsha Stepanek of Business Week refers to this technique as similar to redlining that courts and legislatures have banned when used as a tool by the banking and mortgage industry. But since we will not have access to our “status,” we are not likely to know whether we are being discriminated against. It would reinforce exclusive group membership and social status. If used by communication or information companies, it could increase the inequality of access to information.

Advertisers pay for access to consumer (and voter) information. Once we are placed into a group by a data mining company using neural net technology, we would be unable to challenge our “score”.

Perhaps the best we can hope for is to limit the amount of time transaction data may be stored. But how likely would it be that we could enforce that? It is also frightening that the U.S. Attorney General wants even more sharing between the U.S. and foreign governments. I feel this boils down to a decision based on economics, and bodes dangerous for our social welfare.