A number of questions arise per my previous post on the Netsweeper issue at Toronto Public Library and generally, some of them specific to the Netsweeper issue, and some of them more general. I’ll lay out some of them below — they’re not the only issues, but they’re the ones on my mind at this time.

Is Netsweeper Usage Congruent with Library Values?

Overseas conduct of tech companies in countries with differing levels of legal and practical support for freedom of speech can be a complicated issue; I would not advocate that libraries should cease to do business with a company simply because it operates overseas in countries where active repression of free speech occurs and follows the laws of the jurisdictions where it operates, as this strikes me as both extreme and impractical.

However, I think the Netsweeper issue is categorically different from areas such as the censoring of search results or content takedowns. Based on the recent Citizen Lab report on Bahrain and their previous reports on company conduct, this appears to be a company that actively seeks out the business of low-level Internet censorship overseas, and is assisted by various levels of Canadian government in doing so.

In an ideal world, libraries would not do business with such a company; I recognize the world of library technology usage is often very far from ideal, but also think this is a somewhat less abstract issue than the choice of proprietary software vs. open source, the use of and dependence on outside vendors, or the relative merits of onsite deployment vs. the cloud.

What Can Libraries Do?

At minimum, I think the Netsweeper situation is an argument for libraries having documented policies on their own technology selection and usage at something like the level of consideration and nuance that they have for materials selection and customer Internet use.

Technology choices are not neutral, and organizations such as libraries that align themselves with principles of intellectual freedom should hold themselves to a high standard in dealing with technology that has censorious applications.

What Are the Alternatives?

Over the long term, I believe the only way for libraries to deploy Internet filtering in a way that is congruent with principles of intellectual freedom and other library values needs to meet (at minimum) the following criteria:

  • libraries should use products whose lists of blocked domains in various categories and otherwise are publicly available for audit and scrutiny (Netsweeper, like many other Internet filtering providers, keeps their lists secret).
  • libraries should deploy filtering minimally, only to the extent necessary to comply with legislation and reasonable policy about Internet access in public spaces.
  • libraries should avoid purchasing their filtering software from companies who do business overseas actively facilitating government repression of free speech.

It would be possible for all three of those criteria to be met by a commercial vendor, though I don’t know enough about the commercial filtering software landscape to say if one exists.

My (probably predictable) preference would be for libraries to bend some of their resources towards supporting the development and maintenance of open source filtering software with both public blocklists and examinable code.