This post is mostly a reaction to Adam Thierer’s recent article titled “A Section 230 for the ‘Makers Movement,” although in the grand tradition of reaction pieces it also uses the article as an excuse to write about a few things I’ve been thinking about lately.  It will probably make a bit more sense (or at least be easier to point out what I’m missing) if you read his piece first, although I’ll do my best to summarize it along the way.

The “230″ in his title is Section 230 of the Communications Decency Act, which gives online platforms wide ranging immunity from responsibility for the things their users post on said platforms.  Many people consider it one of, if not the, most important foundational laws for the growth of the internet.  Essentially, this is because it allowed online platforms to let crazy people they don’t know post things for the entire internet to read.  It is the reason, for example, that tumblr isn’t going to do a legal review before letting me use tumblr to post this blog post.  They don’t care what I write because no one can sue them for my degenerate libelous crackpottery.

What Does This Have To Do With “Makers”?

Before Section 230, traditional law treated publishers as responsible for the things that they published.  This made sense when your mental image of a publisher is the New York Times.  The New York Times is exerting a lot of control and review over what gets published in its pages, so it is reasonable to hold them accountable if one of their writers does something crazy.  However, tumblr (or twitter, or facebook, or whatever) doesn’t have the same relationship with its “writers.”  It does not make sense to hold them responsible for what people publish on their platform because they do not exercise any control over that content.  The internet meant that a lot more people were writing for the public and that the old mental model for dealing with that broke down.

You can draw a fairly convincing parallel to the maker movement.  A collection of technologies is allowing more and more people to not only make physical things, but distribute those physical things around the world.  Regulatory models relating to physical things that assume that there are a limited number of people or companies manufacturing and distributing physical objects don’t necessarily apply.  The nature of the new players involved means that traditional legal assumptions may not apply.  That can result in unjust law and crippling (unintentional) inefficiencies.

If that thesis is correct, it is very tempting to look to the internet model - and Section 230 in particular - for the best ways to address this disruption.

Extending Immunity - The Uber Model vs. the Collaborative Model

Obviously talk of immunity for the technologies and platforms that make the maker movement work is highly interesting to me professionally.  After all, my day job is at just one of those platforms.  Similarly, my not-day job is on the board of the Open Source Hardware Association and there are some people who worry about the legal ramifications of releasing a swarm of open source physical objects out into the world.

However, I do have a reflexive concern about any call for blanket immunities for new technologies.  It may not be fair, but some of those are grounded in what might be thought of as the Uber approach to policy.

Uber was faced with a thicket of taxi regulation that threatened their business model. Their initial response was essentially to dismiss those regulations and bulldoze their way through to the market.

For many people this was a highly popular approach, made all the more popular because many of the regulations that were being bulldozed were imposed by government agencies that were captured by the taxi industry they regulated.  The regulations appeared to benefit the taxi industry much more than taxi customers, so many Uber customers quickly sided with Uber over the regulators and did not shed a tear for the lost regulations.

This approach has its appeal.  It can work quickly and radically transform a ossified industry in pro-consumer ways.  But it can also create problems.  One of those problems is that it can alienate potential partners.  But the bigger problem is that the process of bulldozing “bad” regulations can also result in destroying legitimate regulations that have to do with things like encouraging accessibility and protecting users.  This collateral damage isn’t necessarily the intent, but can be very real for users who rely on the regulatory safety net.

There are, of course, alternative models. 

The most obvious is to go back to the concerns that (at least nominally) motivated the regulation in the first place.  Instead of simply saying “this regulation should not apply,” propose an alternative regulatory scheme that achieves the same end goal without whatever negative barriers the existing scheme is creating.  Naturally this assumes that the original goals still make sense, that there is a way to address them without skunking the entire operation, and that someone on the regulator side is interested in having the conversation.  Admittedly, this can also be an excruciatingly slow process.  If it is done well it can leave everyone better off - but that “if” is a big one.

The point of this diversion is mostly to say that I’m a bit wary of simply exempting wide swaths of society from generally applicable rules without some fairly substantial evidence that doing so is necessary and one of the few real options to achieve some legitimate goal.  That’s not to say that it can never make sense - I think Section 230 worked out pretty well - but rather that doing so should usually require a fairly serious consideration of what is lost when granting such an exemption.  

It also means that when thinking about Adam’s proposal the details matter. So let’s get to them.

Models of Immunity

Adam proposes three potential models for immunity: Section 230, firearm manufacturer immunity, and immunity for vaccine manufacturers.

I won’t spend any more time on the 230 stuff.  The firearm manufacturer immunity is interesting.  I’ll be the first to admit that I don’t know a lot about it beyond the fact that Congress passed a law making it hard/impossible to sue firearm manufacturers for tort liability on the theory that they make inherently unsafe products.

Gun politics aside (which is basically impossible in the context of a law involving guns in the US Congress), I’m a bit skeptical that this is a great model for guns or for makers.  One way to think of the purpose of tort liability is to find a way to push an industry towards creating safer products without strangling the industry with regulation.  In that context, I worry that a blanket exception makes it too easy to not move towards safer products in the long run - and “safer products” does not have to mean ‘no more guns’. This is not to say that all tort liability is good.  But I do recognize that it can have a useful role in accelerating the spread of safer products.  The drag that creates for manufacturers can be worthwhile.

The vaccine fund is another interesting example, but one that Adam admits is hard to apply directly here.  The idea is that some sort of government pool of money stands in the place of vaccine manufacturers in cases where vaccines hurt people.  Because everyone benefits from vaccines, in the rare cases when someone is injured by a vaccine it makes sense for everyone to help compensate them.  Removing liability also makes it more likely that companies will step in to manufacture the vaccines and sell them at affordable prices.  The problem when transposing the model to making is that can be quite hard to calibrate the size of the fund or who should be able to hide behind it without more information about the real types of lawsuit threats maker companies are facing.  For better or worse, we just don’t have the data yet.

Areas of Coverage

“Maker stuff” is a crazily broad category, and it would probably be a horrible idea to create some sort of blanket immunity that covers all of it.  Different technologies are different, and the challenges they may or may not present should be dealt with on a more granular level than “it allows people to make and distribute things.”  Easily recognizing this, Adam breaks it down into three categories.


The proposed immunity is for open robot creators if the robot is under the control of the user or software written by a third party.  The idea is that putting a robot out into the world should not make you liable for every crazy thing that someone might do with the robot.

While I like this theory, I’m not convinced (yet) that existing tort liability is inadequate in this situation.  In at least some cases tort can take into account things like third party or user modifications.  Blanket liability exemptions are a big deal and I’d need a lot of convincing that they are really necessary here.  Imposing liability in nonprofessional makers creates a problem, but it can also bring some benefits.

3D Printing

I obviously have a lot of conflicts on this one.  There would be a clear short term (at least) professional benefit to me if all 3D printing platforms were categorically immune from any claim related to what users do on the site.

That being said, I worry about this sort of exemption as well.  “Many people can use this to make stuff” isn’t necessarily a power specific to 3D printing.  When I think about 3D printing policy, one of my first points is usually to avoid treating 3D printing differently unless you have to.  The fear that 3D printing will be implicated because a user does something stupid or illegal is real, but the answer in those cases is probably to treat it like any other technology.  Creating a special 3D printing carveout should only happen when there is real evidence that there is a 3D printing specific problem that needs addressing.  I’m not sure there is one yet.

Virtual Reality

This section of the post feels more like flag planted than anything else.  It is too early in the history of applied VR to know what may or may not be needed.  Adam seems to view it the same way. I’m glad he raises it, and also glad to see that he doesn’t propose anything specific.

Main Takeaways

I found that this piece made me think a lot about the nature of regulation and liability in the maker context, which is fantastic.  When I approach these questions, I try and start by understanding why the existing rules are there in the first place.  Then I try and formulate an explanation for why a specific technology or application changes the assumptions underlying the regulatory regime.  Only at that point would I start to put forward alternative ways to address the underlying concerns of the motivation (or explain why those concerns are no longer valid) that do not unnecessarily hamper the growth of the new technology.  I’m looking forward to the full length treatment of this issue that Adam promises at the end to see how he approaches it.

Clearing Rights for a 'Non-Infringing' Collection of AI Training Media is Hard

In response to [a number of copyright lawsuits about AI training datasets](… Continue reading