Tech as a Democratizing Force - The Rise of AI Facilitated Activism

A big part of crypto advocacy these days is digesting and responding to 100s of pages of dense proposed regulations before responding to them in comment periods, which give stakeholders an opportunity to have their views taken into consideration before the regs are finalized. Typically, how this process works for institutions is that they (i) have a vested stake or position and then (ii) punt on legwork and hire a law firm with specialists in the subject matter of the regs (securities law, tax law, administrative law etc) to draft nuanced arguments after a labor intensive process of reading and digesting the proposed regulations.

The few crypto advocates that we have are constantly put in a position where they need to respond to multiple overlapping agency calls for comments at any given time, which creates bandwidth constraints. In turn, it has made it difficult to come up with grassroots campaigns to generate broader education and response. However, crypto advocates have still long thought that crypto has immense (largely unrealized) potential to be able to leverage its internet native communities — if we can only get over issues around free riding on advocacy efforts and generational apathy toward political processes viewed as inherently unfair.

Enter technology as a democratizing force in the process – on this occasion, a number of advocacy organizations have banded together to come up with tech-based solutions to overcome barriers to participation — leveraging AI and other tools to help mobilize comments on the Treasury’s proposed crypto broker reporting regs.

Like anything/everything else in crypto, these efforts have sparked backlash. Most immediately the negative feedback has come from lawyers — not surprising as we lawyers are natural haters and many of us are also not immune from the basic CT need to engagement farm. In this case, the loudly voiced concerns parrot potential agency responses that would negatively impact the comment process. No good deed goes unpunished but I think it is worth addressing (and refuting) the criticism. 

Specifically, after the AI tool hit Twitter, some lawyers insinuated (among other things) that the use of AI will negatively impact an agency’s obligations under the APA to review each comment letter individually (as is the current obligation) because it amounts to a DDOS attack of the process. The agencies already typically batch form letters and treat them as a single issue to address – the use of AI might create an argument for them to dismiss the letters based on commonalities. To this, I would argue that it is pretty hard to see a path where agencies operating in bad faith would need an excuse to review and respond less than they already do (court is already the only remedy here). But for agencies that interact with stakeholders in good faith, the system should not be dependent on a base level of 99% apathy.

Drilling down into how this AI tool came about, (i) literal tax experts read the regulations, (ii) discussed them at length with industry stakeholders, (iii) produced tailored arguments and then (iv) devs produced a tool and fed those arguments into a tool capable producing hundreds of varying stances and that lets the user/commenter to choose the specific concerns they have with the proposed regs that relate to their own personal interests — really not so different from hiring someone else to hold the pen to articulate your views / concerns in the context of the regulations. The user is also able to set the tone of the comments out of a variety of options (as well as the strength of the tone) to further personalize & is able to review and polish before submitting — not so different from the current process. The product is a text file which is then fully editable by the user to further tailor as needed prior to submission.

In reality, the tool is not a DDOS attack; it is a means of democratizing the existing process by making commenting/participating more accessible to a broader group of citizens, both from a time-spend and economic perspective. So, net, I would argue that lawyers might have a reason to concern-troll because the robots are taking our jobs (!!) but tools that facilitate broader access to citizens to have their views heard are sorely needed. If the government needs to adapt to constructively process more inputs and concerns from a broader group of impacted constituents, that is a positive thing. 

But before you leave — let me use this opportunity to ask that you please consider submitting a comment if you haven’t already (whether AI assisted or made from scratch) before the October 30 deadline.

Try the AI tool here to submit your own comment:

Comments are available for review here:

Leave your comment...

If anyone uses the tool, let me know how it went for you - there is a team still working on improvements.

worked well for me!

It's really cool to hear how the bot was made using various stakeholder arguments. The tool itself was very easy to use and produced a thoughtful & strong response (in less than 1 min). Big shoutout to the whole LeXpunK Army and their efforts to push this space forward.

I used it - worked great! It was shockingly easy to generate the content for submission.

Thanks to lawyer frens @MeatEsq and @larryflorio for letting me bounce this post off of you and for the helpful edits.