New adventures in WiFi

by Maria on September 18, 2003

Hotspots are multiplying all over the place, not just in Stateside Starbucks’, but even along the Paris metro. The only time I’ve used wifi so far was at CFP 2003 where it came in extremely handy for blogging the event. But think of it; free internet, wherever you go – how great is that going to be?

Jonathan Ezor of the Touro Law Center in Huntington, NY has written a piece on three potential problems of publicly available WiFi hotspots; misuse of anonymity, free-riding, and liability of providers. (He actually wrote it in June, I just stumbled upon it this morning.)

The first problem arises from the fact that publicly available wifi hotspots could do away with the need for users to register or identify themselves in some way, tying their computer to a personal identity in meatspace. In some set-ups, users of hotspots will be able to act anonymously, making detection of abuse (DOS or other computer related crime, spam, harassment, etc.) much, much harder. This would certainly be a problem for law enforcement. However, I think Ezor over-estimates current traceability on the internet minus the use of wifi, as this piece by Richard Clayton shows.

As things stand, the G8 Lyon working group on hi-tech crime has been working against anonymity in the communications infrastructure for several years now. Mostly, they’ve been worried about free traditional internet services and pay-as-you-go mobile phones, but you could see how wifi could become a real concern. I’m no expert on the tech aspects of this, (and thoughts/expertise are very welcome) but it seems to me that the software could be set up to make some sort of registration process mandatory. As wifi develops and the various pilot projects conclude, we may see more registration requirements, particularly for hotspots provided by public agencies or public/private partnerships. People can always provide false information, but that’s no different from users’ existing relationships with ISPs. And ultimately, we’ll see the law-abiding signing up, and the bad actors, as always, finding ways to escape monitoring and detection.

Free-riding seems to present another teething problem. What to do about companies or users who hog bandwidth and slow down or even stop publicly provided network access? This raises an interesting question – to what extent is a wifi hotspot a public good? It’s neither purely indivisible nor is it, depending on registration requirements and network monitoring capability, entirely non-excludable. Bandwidth is finite, so it is to some extent a rivalrous good. It all depends on who owns the system and how it’s set up of course. But, over time, the trade off between network efficiency and user convenience may also tend toward registration requirements which will provide a means to prevent bandwidth hogs doing their thing.

Ezor calls for a liability safe harbour for wifi hotspot providers, pointing to existing (US) liability exemptions from libel claims. If only libel claims were the biggest problem communications service providers faced. What is really needed is liability protection against IP and copyright claims. Under EU legislation (specifically, Article 12 on ‘mere conduit’ of the E-Commerce Directive) providers of communication services should be exempt from liability for the information they provide. But not only are content owners taking legal action against various European CSPs who could not possibly be expected to monitor every piece of information on their networks, but even this protection is now to be removed altogether by the draft IPR Enforcement Directive. So how about a future in which the RIAA sues not only grandfathers and 12 year old girls, but also local councils, hospitals, schools, and, heaven forbid, mass transit systems?

At least in the early days of wifi, the technology probably will be used by early-adopting criminals (amongst others). Forget infrastructure and rollout costs. Liability, risk, and the expense or impossibility of insuring against them are the most likely candidates to smother wifi at birth.



Cory Doctorow 09.19.03 at 3:21 pm

Anonymity is a benefit, not a bug, from where I sit. I’ve run open APs in Toronto and San Francisco for four+ years now, and one of my main reasons for doing so is to allow for anonymous speech, for the same reason that the framers of the Constitution supported anonymity.

As to free-riders: not only have I never had this problem (contention for my bandwidth), but I’ve never heard of anyone having this problem. My APs accomodate 10 users, my DSL lines deliver 1.5Mbs and most websites deliver < 60ks in throughput. In that scenario, contention can NEVER occur -- unless you have one guy downloading an enormous file from Akamai or Bittorrent. It's an edge case, it doesn't really occur in the field, it's not a problem that needs solving, unless you've got the economist's irrational, superstitious dread of free-riding. (In any event, the NoCat and other free-software open AP projects manage bandwidth contention either through fair apportionment or prioritization, and includes a login system) Finally, is it really to an AP operator's benefit to implicitly assert that she has the right, ability or duty to control her users' actions? Safe harbors are nice, but I'd rather get common carrier status: I take all traffic, I don't have the right, ability or duty to monitor it, and I send it on. (Really, is anonymity in WiFi more socially corrosive than anonymity in prepaid mobile phones? How about anonymity in message-board posting?)


james crabtree 09.24.03 at 11:16 am

hey maria – nice blog. just thought i’d pop by and have a look. Me and the 100,000 others… :-)

keep in touch. jc

Comments on this entry are closed.