Building an open web that protects us from harm

We live in a world where right-wing nationalism is on the rise and many governments, including the incoming Trump administration, are promising mass deportations. Trump in particular has discussed building camps as part of mass deportations. This question used to feel more hypothetical than it does today.

Faced with this reality, it’s worth asking: who would stand by you if this kind of authoritarianism took hold in your life?

You can break allyship down into several key areas of life:

• Who in your personal life is an ally? (Your friends, acquaintances, and extended family.)
• Who in your professional life is an ally? (People you work with, people in partner organizations, and your industry.)
• Who in civic life is an ally? (Your representatives, government workers, individual members of law enforcement, healthcare workers, and so on.)
• Which service providers are allies? (The people you depend on for goods and services — including stores, delivery services, and internet services.)

And in turn, can be broken down further:

• Who will actively help you evade an authoritarian regime?
• Who will refuse to collaborate with a regime’s demands?

These two things are different. There’s also a third option — non-collaboration but non-refusal — which I would argue does not constitute allyship at all. This might look like passively complying with authoritarian demands when legally compelled, without taking steps to resist or protect the vulnerable. While this might not seem overtly harmful, it leaves those at risk exposed. As Naomi Shulman points out, the most dangerous complicity often comes from those who quietly comply. Nice people made the best Nazis.

For the remainder of this post, I will focus on the roles of internet service vendors and protocol authors in shaping allyship and resisting authoritarianism.

For these groups, refusing to collaborate means that you’re not capitulating to active demands by an authoritarian regime, but you might not be actively considering how to help people who are vulnerable. The people who are actively helping, on the other hand, are actively considering how to prevent someone from being tracked, identified, and rounded up by a regime, and are putting preventative measures in place. (These might include implementing encryption at rest, minimizing data collection, and ensuring anonymity in user interactions.)

If we consider an employer, refusing to collaborate means that you won’t actively hand over someone’s details on request. Actively helping might mean aiding someone in hiding or escaping to another jurisdiction.

These questions of allyship apply not just to individuals and organizations, but also to the systems we design and the technologies we champion. Those of us who are involved in movements to liberate social software from centralized corporations need to consider our roles. Is decentralization enough? Should we be allies? What kind of allies?

This responsibility extends beyond individual actions to the frameworks we build and the partnerships we form within open ecosystems. While building an open protocol that makes all content public and allows indefinite tracking of user activity without consent may not amount to collusion, it is also far from allyship. Partnering with companies that collaborate with an authoritarian regime, for example by removing support for specific vulnerable communities and enabling the spread of hate speech, may also not constitute allyship. Even if it furthers your immediate stated technical and business goals to have that partner on board, it may undermine your stated social goals. Short-term compromises for technical or business gains may seem pragmatic but risk undermining the ethics that underpin open and decentralized systems.

Obviously, the point of an open protocol is that anyone can use it. But we should avoid enabling entities that collude with authoritarian regimes to become significant contributors to or influencers of open protocols and platforms. While open protocols can be used by anyone, we must distinguish between passive use and active collaboration. Enabling authoritarian-aligned entities to shape the direction or governance of these protocols undermines their potential for liberation.

In light of Mark Zuckerberg’s clear acquiescence to the incoming Trump administration (for example by rolling back DEI, allowing hate speech, and making a series of bizarre statements designed to placate Trump himself), I now believe Threads should not be allowed to be an active collaborator to open protocols unless it can attest that it will not collude, and that it will protect vulnerable groups using its platforms from harm. I also think Bluesky’s AT Protocol decision to make content and user blocks completely open and discoverable should be revisited. I also believe there should be an ethical bill of rights for users on open social media protocols that authors should sign, which includes the right to privacy, freedom from surveillance, safeguards against hate speech, and strong protections for vulnerable communities.

As builders, users, and advocates of open systems, we must demand transparency, accountability, and ethical commitments from all contributors to open protocols. Without these safeguards, we risk creating tools that enable oppression rather than resisting it. Allyship demands more than neutrality — it demands action.

https://werd.io/2025/building-an-open-web-that-protects-us-from-harm