Online Dignity and the Cost of Normalised Harassment

Online communities have a design flaw baked into their reward systems. Outrage gets engagement. Harassment scales effortlessly. A single thoughtful critique gets 3 replies; a pile-on gets 50. The algorithms learned this before we did, and now most of us just accept that digital spaces are fundamentally meaner than face-to-face ones.

This isn't accidental. It's structural.

When we normalise harassment, we're not just being rude to individuals. We're filtering out entire groups of people: those with anxiety disorders, neurodivergent people who process rejection differently, people in precarious housing situations, disabled people who cannot afford to absorb trauma, and communities already targeted by systemic discrimination. We're selecting for thick skin, which is another way of saying we're selecting for privilege.

The worst part? We've convinced ourselves this is fine. "If you can't handle criticism, don't post online." "Toughen up." "That's just how the internet is."

Except it doesn't have to be.

Why Critique and Respect Are Not Opposites

The Humane Interaction Standard starts from a simple observation: you can dismantle someone's work without dismantling their personhood.

This is not radical. It's basic. A mathematician can tell you your proof is wrong without calling you stupid. A designer can say your colour palette is broken without attacking your aesthetic taste as an insult to your character. A developer can argue that your architecture won't scale without suggesting you should never code again.

We all know how to do this in academic settings. Peer review exists specifically because the research community understood—centuries ago—that rigorous critique requires separation. You critique the manuscript, not the author's mother.

Yet in open source, in creative communities, in AI discourse, we've somehow lost this boundary.

The HIS doesn't ban criticism. It doesn't ask people to be nice. It asks communities to enforce a basic rule: critique the work; don't weaponise personal information or irrelevance against the creator.

This is about structural integrity, not tone policing.

The "Tool Choice" Problem

One specific place where this matters urgently: the AI conversation.

There's a lot of justified anger about AI ethics. The environmental cost is real. Consent for training data is a real problem. Labor displacement is real. But somewhere along the way, that legitimate critique got tangled with personal attacks on individual creators, tool developers, and researchers who choose to work with these systems.

"You used AI" becomes "You're unethical" becomes "You don't deserve a platform" becomes "Let's find where you live."

The HIS makes one thing clear: tool choice is not a moral disqualification for being treated as human.

This matters for AI practitioners (yes, even when you disagree with them). It also matters for AI critics. If someone critiques an AI system with civility, they shouldn't be derided as a corporate shill or a Luddite either.

The work can be wrong. The methodology can be flawed. The impact can be harmful. But none of that justifies harassment.

Harassment Is Not the Same as Criticism

This is where enforcement gets tricky. Communities adopting the HIS need clear language around what actually constitutes harassment.

One harsh critique: Not harassment.

Five separate threads, all tagging the creator after they've asked you to stop: Harassment.

Pointing out serious flaws in someone's work: Not harassment.

Coordinating with others to overwhelm their mentions, replies, and DMs: Harassment.

The pattern matters. The intent matters. The scale matters. The persistence after being asked to stop matters.

Most moderators know this intuitively, but they rarely formalise it. HIS does. It gives you language to say, "You're not wrong, but this is harassment," without being wishy-washy about actual problems.

Why This Matters for Disabled and Neurodiverse People

Here's something the standard documentation doesn't emphasize enough: online harassment has a specific cost profile.

A neurotypical person with secure housing and low anxiety might see a pile-on, feel bad for a day, and move on. A neurodivergent person with a processing disorder might spiral for weeks. Someone in a precarious situation might lose focus on work or study. Someone with chronic illness might have a flare triggered by stress.

We don't usually measure the cost this way. We just measure engagement metrics.

Adopting the HIS isn't charity toward disabled people. It's recognising that when you normalise harassment, you're imposing a hidden tax on people already living with invisible disabilities. You're filtering out perspectives because those perspectives come from bodies and brains that can't afford to be treated this way.

The standard applies equally. Everyone gets the same baseline: your work can be critiqued harshly; your personhood is not up for grabs.

Enforcement Without Becoming Authoritarian

This is the hard part. How do you enforce humanity without becoming a surveillance-state moderator?

The HIS model is proportional:

• First violation: warning and explanation
• Repeated or severe violation: content removal, mute, ban

No zero-tolerance theatre. No performative cruelty dressed up as justice. Just clear expectations and graduated response.

The trick is trusting your moderators (and holding them accountable). They need to understand the difference between "harsh feedback" and "harassment." They need to feel empowered to warn without immediately banning. They need to know that culture change is slower than spectacle.

This requires training and time. Most communities don't invest in either.

A Small Thing That Matters

Adopting a standard like HIS won't fix the internet. But it does a few concrete things:

1. It makes expectations explicit. New members know what the space tolerates. Chronic harassers know where the line is (and whether they want to stay).

2. It gives moderators language and authority. Instead of vague appeals to "be nice," mods can say, "This violates section 3.2 of our standard. Here's why."

3. It signals values. If you adopt HIS, you're saying: "We believe critique and dignity are compatible. Tool choice isn't a justification for abuse. Harassment is a pattern, not a feeling."

4. It's CC0. You can copy it, fork it, adapt it to your community. You don't need permission.

The standard itself is straightforward enough to fit on a page. The philosophy is simpler still. But getting communities to actually enforce it, to actually change culture around harassment—that's slower work.

Worth doing though.

Beyond the Standard

The Humane Interaction Standard is a starting point, not a solution.

What actually changes culture is consistent enforcement. A moderator who actually removes content violating the standard. A community that reports harassment instead of ignoring it. A culture that separates "your work has problems" from "you're a bad person."

This is especially important in open-source communities, AI discourse, and creative spaces where power dynamics are already skewed. The creator is usually more vulnerable than the critic. The minority perspective is usually more vulnerable than the consensus. The person with fewer connections is more vulnerable than the person with networks.

The standard doesn't remove these asymmetries. But it does say: we're not going to weaponise them.

That's a small thing. And it matters.

The Humane Interaction Standard (HIS) is released into the public domain under CC0 1.0 Universal. You can adopt it, modify it, or use it as a basis for your own community standards. For the full standard, see the repository.