Salon Magazine
Thursday, July 31, 1997

Ratings Today, Censorship Tomorrow

The Net industry is rushing to embrace ratings systems for the Web. The technology will help parents keep their kids away from porn. It can also help anyone censor anything.

by Joseph D. Lasica, lasica@pacbell.net

A few years from now, when we look back at what crippled the Internet as a global forum for the free exchange of information, at least we'll know it was done with the best of intentions.

Who, after all, could oppose Internet ratings if they create a "family-friendly" online world?

And so, to make the Net safer for kids and to avert government regulation, the Internet brain trust has banded together to push rating, filtering, and labeling technology -- a private-sector techno-fix to cleaning up the Net. President Clinton has signed on and has used his bully pulpit to jawbone companies that were wavering on the issue. And the news media have covered the president's initiative with the gusto of a pep rally.

With all this firepower behind them, ratings are coming to a Web site near you -- in fact, to all Web sites, if proponents have their way. And a panoply of would-be censors -- from foreign despots to home-grown zealots and pandering politicians -- couldn't be happier.

"What's happening now is a move toward the privatizing of censorship", says David Sobel, legal counsel for the Electronic Privacy Information Center (EPIC). "It's likely to destroy the Internet as it's existed up till now."

There are a great many ironies here, but the greatest irony is that the censorship will be self-imposed -- we're doing it for the sake of family, parents, children. In truth, Internet ratings are being driven by the changing business interests of the major players involved.

In the last go-around over muzzling the Internet, Net users, the computer industry, the media, and civil liberties groups all united against the government's Communications Decency Act -- which the Supreme Court buried last month. This time around, the lineup is a lot more lopsided.

On one side: the U.S. government, the high-tech industry, most major media outlets, and a vocal cast of parents' organizations, child-safety advocates, and anti-obscenity groups.

On the other: the American Civil Liberties Union, EPIC, the American Library Association, a smattering of university scholars, and that guy over there waving the "No ratings" sign.

Why have the software companies and Internet firms gone over to the other side? Certainly, they're spooked by the specter of Congress passing a "son of CDA" bill. But it goes beyond that.

Internet ratings dovetail nicely with big business's desire to make the Internet safe for God, apple pie, and commercialism. The "dark side" of the Net -- hackers, foreigners, political extremists, geeks, "phreaks", porn purveyors, hate groups, people who SHOUT IN ALL CAPS AND USE EXCLAMATION MARKS!!! -- will largely be banished to an unrated no-man's land where browsers and search engines fear to tread.

So it was no surprise that the invitation list to the Internet summit at the White House on July 16 bore names like Netscape, America Online, and IBM rather than names like geekboy or cybergrrrl. At the meeting, President Clinton announced a "parental empowerment" initiative that would give parents the tools to shield children from obscenity, violence, and antisocial messages on the Net. Although every idea on the table is software-based, the administration couldn't resist dubbing the plan the "E-chip", a cousin of television's V-chip, which will block unsuitable programming.

"We need to encourage every Internet site, whether or not it has material harmful to minors, to rate its contents ... to help ensure that our children do not end up in the red-light districts of cyberspace", Clinton said.

And the assembled captains of industry obliged. Netscape indicated it would support Internet ratings in its next browser, meaning that more than 90 percent of all browsers will support Internet ratings. (Microsoft's Internet Explorer 3.0 already includes ratings as an option for parents to turn on.) The search engines Yahoo, Lycos, and Excite also fell into line, pledging to ask for self-rated content labels for all Web sites on their directories.

There's just one problem with all this: "Childproofing" the Net by labeling content is likely to be an unmitigated disaster for adults.

"Unfortunately, a lot of people think we need to knock down everything to the common denominator of this mythical 6-year-old who surfs the Net", says Sobel of EPIC. "If this trend continues, the Internet is not going to be the open forum of ideas that it has been."

"These efforts to rate the Net result from a real misunderstanding of what the Internet is all about", says Jaron Lanier, a visiting scholar at Columbia University and computer scientist who coined the term "virtual reality". "The Internet is not just another medium choice, like television or the movies. It's the future of all communication that's not face to face. To say that we're going to rate all communication is a criminal idea."

"This will have a devastating effect on free speech all over the world -- and at home", declares Lawrence Lessig, a professor of law at Harvard Law School and one of the foremost authorities on Internet ratings. "To my mind, PICS is the devil."

PICS, or Platform for Internet Content Selection, is the labeling language developed at MIT that allows Web pages to be rated and screened out. In theory, dozens of rating systems could be used with PICS technology; the Christian Coalition could rate sites on their godliness, and the ACLU could rate them on their friendliness to free speech. But in practice, only three groups have devised actual rating systems based on PICS: SafeSurf, Safe For Kids, and the de facto industry leader, RSAC.

RSAC is short for the Recreational Software Advisory Council, and if that sounds like a strange term for a body wielding such enormous power over free-speech issues, it may be because the group was originally set up in 1995 to rate video games. In April 1996, its mission was expanded to devise a rating system for the Net. The nonprofit group in Cambridge, Mass. -- just down the road from MIT -- is now backed by IBM, Dell, Disney, CompuServe, Microsoft, and leading media companies.

For site operators, RSAC's Internet rating system (RSACi) works like this: You connect to the RSACi Web site and fill out a form rating your site for sex, nudity, violence, and offensive language. Then you're assigned a tag and slap it into your Web page's HTML code. The tag is invisible to anyone looking at your Web page but can be read by PICS-enabled browsers, search engines and "censorware" software products like Net Nanny and Cyber Patrol.

Under this rating system, the end user can set a tolerance level of 0-4 for each content category. You could allow "moderate expletives" (level 2) or screen out "strong language" (level 3). You could permit "clothed sexual touching" or draw the line at "passionate kissing". It's all intended to make for an idyllic, family-friendly, Frank Capra kind of browsing experience.

But the Net has begun to buzz with critiques of PICS, RSACi, and Net rating systems. They've been derided as parochial, inflexible, culturally biased to reflect the prejudices of those doing the rating, unable to distinguish between fiction and nonfiction and more appropriate to computer games than to text and complex ideas.

Moreover, such a clunky Web-based system seems irrelevant to a large chunk of cyberspace. Ann Beeson, the ACLU staff attorney who helped bring down the CDA in ACLU vs. Reno, observes: "Everybody thinks of the Internet as the Web, but they don't think of e-mail, or Internet Relay Chat, or the hundreds of Usenet newsgroups with no person in charge, or bulletin boards and conference threads. How do you rate those?"

But let's put aside all of these criticisms for a moment. Even if RSACi and all the other PICS-based Internet rating systems worked perfectly, they would still suffer from one monstrous flaw: The user may not be the one making the decision on what material is screened out.

"The problem", Lessig points out, "is that the filter can be imposed at the level of the individual user, the corporation, the proxy server, the Internet Service Provider or the national government. This is disastrous, because you can have invisible filtering done at any level of the distribution chain."

That's the essential difference between filtering and censoring: Who decides what you can see?


People talk about how PICS empowers the user. But PICS is two-sided: It supports self-labeling and "third-party rating". On one side, people can rate their own documents using whatever rating system they want. On the other side, rating authorities can also rate other people's content -- just as Web censorware now works -- and distribute their ratings online. Or, if they have the power or the legal authority, they can just cut off access to sites they don't like. This is not content identification, it's content control.

"We're complacent about the nature of the Net", Lessig says. "We think it's unregulatable and that it can be no other way. Well, PICS changes the architecture of the Internet into an extremely regulatable structure. You then have a versatile and robust censorship tool, not just for parents but for censors everywhere. It will allow China and Singapore to clean up the Net. It will let companies control what their employees can see. It makes it easy for school administrators to prevent students from viewing controversial sites."

This is no exercise in academic conjecture. Already, Australia, Japan, and Dubai are weighing labeling plans to muzzle the Net. And censorware programs are already in use in some public library systems and public school systems.

Soon, third-party intermediaries -- from employers, libraries, universities, and access providers to the Internet cafe down the street -- may substitute their judgment for yours. And that's the most subversive part of this: You may never know that a particular article or idea or site even existed.

Indeed, censorware products are already blocking access to political organizations, medical information, and unpopular viewpoints. And chances are that it won't end there. Legislation, boycotts, and pressure tactics may be brought to bear against Internet service providers that refuse to block sites featuring alternative lifestyles, extremist political views or fringe ideas.

We already have a recent precedent. Remember those "advisory" labels on music CDs? Andy Oram, moderator of the Cyber-rights mailing list for Computer Professionals for Social Responsibility -- which opposes Internet ratings -- says, "Music labels were supposed to be advisory, but we've already seen albums excluded from stores, and several states passed laws barring teenagers from buying albums with advisory stickers. What started out as a tool for parental empowerment turned into an effective means of censorship."

Already, several Net rating bills have been floated in Congress, ranging from government-mandated labels to criminal penalties for those who mislabel their site. One proposal, the Online Cooperative Publishing Act, was put forward by SafeSurf to ensure that families "may feel secure in their homes from unwanted material".

SafeSurf, which is lobbying mightily to become the rating system for Netscape, goes well beyond RSAC's sex, nudity, violence, and language categories and five levels of access. Instead, it offers nine categories -- including gambling, "glorifying drug use", and "homosexual themes" -- and nine rating levels. (It also screens for "intolerance of another's race, religion, or gender".) The company, which started out in 1995 as a two-person parents' group, just moved into plush offices on Wilshire Boulevard in Los Angeles, backed by large investors.

"A lot of parents we hear from are more concerned with gangs and gambling and neo-Nazis than with Playboy", says company president Wendy Simpson, who claims SafeSurf has members in every nation on the planet.

Simpson notes, correctly, that PICS filtering is value-neutral -- and indeed one can imagine many uses beyond screening for "adult" content. She says the Simon Wiesenthal Center has contacted SafeSurf about devising a rating system to screen out Nazi hate groups and sites that glorify the Holocaust.

In short, Simpson doesn't think there are any free-speech issues involved at all. "The day is going to come when it'll just seem to be the natural thing to do, to rate your own site. To tell you the truth -- and I told President Clinton this -- I don't think there is resistance [to Internet ratings]. I think there's a lack of public awareness."

Stephen Balkam, executive director of RSAC, also dismisses free-speech concerns and the prospect of censorship by repressive regimes: "It's unfortunate that the Chinese have moved in a different direction from the rest of the world. But you can hardly blame technology for what these governments decide to do. It's like saying electricity is to blame for electrocutions."

For its part, Netscape is even more sanguine about the implications for global censorship. Peter Harter, the company's global public policy counsel, says Netscape is "value neutral" on the subject of censorship.

"Netscape has responsibilities as a global information company", he says. "Other countries don't have a First Amendment, and we don't believe it's our right to force our values down the throats of other cultures and countries. If Singapore or China or other nondemocratic countries choose to set up massive filters to restrict the information flowing into their country, it's their right."

Why not make the issue moot by developing a technology that can be used only at the desktop user level? Says Balkam: "That ignores what goes on in the workplace. I can see the value of a company setting a Level 2 for nudity as a way to prevent employees from spending two hours a day looking at porn sites."

It seems that snuffing out political dissent on the Net is just the price the developing world will have to pay in order for businesses to crack down on employees peeking at nude pictures.

As if all this weren't controversy enough, RSAC is about to wade into the uncharted territory of determining which publications qualify as "news" sites on the Internet. The board has already approved an "RSACnews" rating that would exempt news organizations from having to rate their own Web pages. Users could then click on a separate lever in their browser to decide if they want to receive news or not. Microsoft has given preliminary acceptance of the news category for Internet Explorer 4.0, Balkam says.

Such an exemption is RSAC's tacit admission that ratings are unworkable for such things as news coverage. How, after all, do you rate a story about the murder of Gianni Versace, or Bosnia war victims, or the Oklahoma City bombing? And how do you decide where news itself begins and ends?

RSAC and an adjunct advisory board, the Internet Content Coalition -- made up of representatives from the Wall Street Journal, MSNBC, Sony, Prodigy, Playboy, Ziff-Davis, Ad Week, CNET, and Wired, among others -- intend to judge which news sites qualify as "legitimate", based on criteria now being drafted. Thus, mainstream news organizations will sit in judgment of small, alternative, activist publications.

Will sites specializing in soft news, analysis or features qualify for the news rating -- sites like Slate, Suck or Salon? Balkam thinks not: "We'll follow a similar line to television ratings, where soft news items are rated."

As ratings systems are installed across the Web, the most important question to ask at each step of the way will be, whose hands are on the controls?

Paul Resnick, an associate professor at the University of Michigan School of Information who helped devise the PICS standard, sees a future in which control is widely distributed among hundreds of different rating systems: "We'll be able to share one Internet but see the network, and each other, through our own lenses."

But are they lenses -- or blinders? It doesn't take a wild leap of the imagination to see where all this is leading. Islamic countries will block out Western values on the Net. Iraq and Iran will censor their dissidents. Israel will ban sites sympathetic to the Palestinians.

Virtual reality pioneer Lanier, who considers PICS more dangerous than the CDA, cautions: "The Internet creates a giant mirror where we see the whole of humanity -- the bad with the good. If you start creating these narrow rating channels by precensoring opinions and ideas before you've even been exposed to them, then our lives will be dimmed and narrower and the sky a little less bright."

In the ferocious stampede to protect children in cyberspace, many people who should know better are willing to balkanize and fragment the Internet. If we succumb to the hype of ratings and labeling, our new lenses will show us only a flattened, one-dimensional view of the online world. Mass-audience corporate Web sites will be spared, but ratings will blind us to many of the quirky, idiosyncratic, vibrant voices that make the Internet so astonishing.

And to what end? Warns Sobel, the EPIC counsel: "This kind of technology will sanitize content to the point that it's even safer and less controversial than the mainstream media."

That should scare us infinitely more than what's out there on the Net today.


Copyright © 1997 by Joseph D. Lasica. All Rights Reserved. Reprinted with permission.