New tests show Web filters don't screen objectionable material fully

By The Associated Press

02.15.01

Printer-friendly page

NEW YORK — Internet filtering software generally fails to block one out of every five sites deemed objectionable, Consumer Reports magazine concludes.

The magazine said filters hadn't improved since it last tested them four years ago. The latest results appear in the March issue, scheduled to hit newsstands Feb. 20.

"Many parents continue to buy these products possibly under the impression that their children are perfectly safe,'' senior editor Jeff Fox said yesterday. "Our results suggest they should not rely solely on filtering software to be a baby-sitter.''

The report was issued as schools and libraries across the country prepare to install filters to protect children from pornography in order to receive federal funding for Web access.

The American Civil Liberties Union and the American Library Association plan court challenges to the new requirements, which take effect later this year.

Fox warned that the new law could imply endorsement from Washington, even though filtering packages still need work.

Using lists of 139 sites known to contain questionable or controversial materials, reviewers at Consumer Reports tested six software packages along with parental control features available to America Online subscribers.

They rated the filters' ability to protect children from "objectionable material,'' including sexual content and the promotion of crime, bigotry, violence, tobacco and drugs.

Consumer Reports found AOL's Young Teen settings "pretty effective,'' failing to block only 14% of objectionable sites. But those settings also blocked 63% of legitimate sites because the filters only allow access to a preapproved list.

AOL's Mature Teen settings do not rely on a preapproved list and instead block sites on a prohibited list. Consumer Reports found the filters let 30% of objectionable sites through.

Consumer Reports said Cyber Snoop and Net Nanny offered poor protection, while Cyber Patrol, Cybersitter 2000, Internet Guard Dog and Norton Internet Security 2001 had fair or good protection.

Filters can differ in effectiveness because of both methods and value judgments — for example, are pictures of aborted fetuses objectionable?

David Fertell, chief executive of Pearl Software Inc., said Consumer Reports had mischaracterized the purpose of the Cyber Snoop software his company makes.

He said the package was designed primarily as a logging system, so parents can monitor how kids use chat rooms and visit Web sites.

Net Nanny spokeswoman Nika Herford said the company made its software and filter lists customizable, and their effectiveness depends on the settings used.

Consumer Reports also checked for how well the filters permitted potentially controversial but legitimate information on such topics as sex education and gun control. With the exception of AOL's Young Teen settings, the packages got good to excellent ratings.

But the magazine's Fox was still troubled. He noted a government drug-control site was blocked by one filter because it named a drug.

Kevin Blakeman, president of Surf Control Inc., said the company never claimed its Cyber Patrol software was 100% effective. "We're not really a replacement for parents, but a helping hand for when they are not there.''

The general findings are in line with those of the National Coalition for the Protection of Children & Families, an anti-pornography group based in Cincinnati.

Jack Samad, the coalition's vice president for Internet safety and communications, said education by parents is more important than any technological solution.

But he said filtering packages do serve a purpose, despite their flaws, just as seat belts in cars are useful even if they don't prevent all injuries.

"Let's reduce the harm to the point we can today,'' Samad said, "and continue to work on better solutions.''