This year I taught my “standard” platform regulation course for something like the twelfth time. Here’s the syllabus. The course has changed a lot over the years.
Some of those changes come from real world developments. The biggest one recently is the expansion of U.S. constitutional law coverage. (To compensate, I dropped some older international human rights materials. But let the record state that those materials exist. Other parts of the world noticed some key questions years before the U.S. did!) Other changes come from my own thinking about the law or how to teach it. I’ve experimented with a lot of organizational approaches. They all have pros and cons. (Should all of Section 230 be taught together? Or should all “must-carry” law be taught together, so some 230 cases go with the First Amendment issues from Moody? Should the DMCA, CDA, and DSA be on adjacent days for comparison?)
This year I experimented with emphasizing merits liability questions more. That fits with a framework that I used in class and increasingly use in explaining platform issues outside of class.
For many questions about online speech, we need to consider (1) merits liability, (2) statutory immunities, and (3) the First Amendment. Sometimes you don’t need to analyze all three to reach your legal conclusion. But you’ll catch a lot of issues by going through the checklist. Consciously breaking out those three questions also helps in sorting out how the three interrelate, as in cases like Anderson v. TikTok.
Some other reasons to spend more time on merits liability questions:
- We now have a Supreme Court ruling about civil liability for platforms, Taamneh.
- Federal criminal laws apply to platforms and are often under-discussed.
- If Section 230 goes away, merits will matter. Relatedly, your expectations about merits liability should shape your predictions and policy preferences about amending 230. (IMO if 230 went away, Americans could litigate tort claims for a very long time without arriving at much legal certainty. There are just too many varying fact patterns and legal standards, and too much incentive for litigants to keep trying.)
- Lots of potential 230 cases are actually resolved on the merits now. I suspect that trend is increasing.
- It’s hard to talk clearly about secondary liability for platforms if students are not clear about primary liability. They need to know what actions by users or platforms could violate the law in the first place.
- The mental state (scienter or mens rea) and bad act (actus reus) elements of many merits claims have analogs in most immunity statutes. Considering those things through both lenses is useful in providing practical legal advice, I think.
- Students aren’t always clear on the idea of elements a plaintiff must prove and defenses a defendant can raise. If I teach them only that, it’s a win for the legal profession.
Here are some more scattered observations about the changes.
The Law of the Horse: I have started every class for years with Larry Lessig’s Law of the Horse article. It is great for reminding students to think about factors other than law, and to consider other tools for solving problems. It also tees up questions about how the law should respond when technological change makes our information environment “work” differently. Is our goal to get back to the legal status quo? Or is that a moment to reconsider the values and priorities that motivated the law in the first place, or should motivate it now?
Semesters versus quarter: I first taught at Duke Law School, and then at Berkeley, both of which are on semester systems. Switching to Stanford’s quarter system was hard. I am constantly eliminating topics, and never feel that anything I cover gets enough time. I don’t understand how subjects like Property possibly get taught in one quarter.
Practitioner-oriented teaching: In the early 2010s I taught with in-house practice as a frame, and I loved it. I do it less today because I haven’t done that work in a while -- though now if there is time I talk about things like communicating with policymakers. I used to cover things like the motivations of particular teams within a platform company (Sales, Product, Eng, Comms, Public Policy, Legal, etc.), and who might be allies or adversaries for particular issues. Some of my exams from this era provided hypothetical scenarios and asked students to communicate their legal advice in one set of slides for the CEO and a separate explanatory email to the GC, since we had discussed tone and framing for these different communications. My favorite exam question from that period started with a “someone in country X has a legal claim and you don’t know the law there” scenario; told the student they only had budget for four hours of outside counsel advice; and asked them what questions they would prioritize. (I kind of love exams.)
Immunities: For speech and content issues, 2010s classes were mostly focused on immunities under the CDA, DMCA, and some non-US laws. The only content topic where I covered the merits of claims in detail was copyright. When I taught with Mark Lemley in 2016, the class was just called “Intermediary Liability.” (That was fun, and also required a lot of disclaimers about our potential biases. Between the two of us, I’m pretty sure we had worked on the majority of the cases we covered.)
Copyright: Copyright used to take up much more of the class, since that gets litigated a lot and can be a make-or-break issue for platforms financially. These days students are less interested in it, and there’s just a lot more non-copyright material to talk about. So it only takes up a day or two, which really undersells its importance. I had to drop DMCA 1201 and 1202 almost completely. (Sniff!) I still try to drive home that, for platform lawyers, copyright may be where their legal advice has the most serious consequences for the client, and in which the most fraught new issues or conflicts may arise. I miss copyright! The politics are hideous, but the doctrine is so much fun.
Privacy: I used to include privacy and surveillance in the course. The part I was most excited about, but never quite got right, was a privacy policy drafting exercise. (Here’s the 2014 version.) This turned out to be too ambitious, even when spread over two and a half classes with (theoretical) ongoing group work outside of class. I still think that done right, this could be a powerful teaching tool. It could be combined with TOS drafting, or drafting a workflow for handling takedown requests or user data requests.
I wanted students to work through (1) drafting a meaningful policy to help users understand the privacy concerns and tradeoffs of a particular product, and (2) identifying in-UI options to give users choices or notice at time-of-sign-up or time-of-use for specific features. I’ve also tried having them (3) advise clients to change particular features, and (4) defend their legal decisions to another student playing the role of an FTC lawyer. And once on an exam I put the students in the shoes of an in-house lawyer, and spelled out that they had the option (and financial means) to (5) quit over the CEO’s choice to do something super sketchy with user data. No one quit.
Things you can’t teach in one day: Teaching EU Right to Be Forgotten to students who have never encountered data protection law is very hard to do in one day, and I stopped trying. Jurisdiction and cross-border content takedown issues are too complex, you can’t communicate the boring doctrine along with the truly interesting issues in one day and I can’t justify spending more class time on it. I usually include the topic anyway. I try to leave them intrigued, aware of pitfalls, and with a mental checklist of things to consider in practice. But I know they also inevitably feel confused about the law. Teaching the DSA in one day is also a fool’s errand, but I try to do that, too. I do a high level overview and then zoom in on just a few Articles.
