Civic Insurance
Civil society: hedge, not opposition
Last week I introduced you to my grandfather, who served the Detroit Polish community with his hardware store. He loved his Manhattans and spoke fondly of the Old Country. He played cards, and — like my father — I never once heard him use a cuss word. Which is why he’d probably turn over in his grave reading about enshittification (Cory Doctorow’s term) and e/ensh (my term) as a description of what’s become of (American) capitalism today — a system that has lost many of the core values he grew up with and practiced in his own life and with his customers.
That piece, Digital Landlords, identified a destructive dynamic at play within capitalism today: the enshittification of both business models (enshittification) and business leadership culture (e/ensh - effective enshittification —> read the piece to learn more!). On one hand we’re baht and rupiah’d to death over subscription fees and false pretenses of value-added, while simultaneously witnessing the deterioration of business leadership, such that people are commodified, relationships become meaningless, and loyalty becomes a one-way extraction.
Okay, so what checks all this?
This essay isn’t about how to fix bad business models and bad leadership — that’s for another day. Rather, I’m focussing on how we can minimise the damage those dynamics create for citizens of the Rest of World, and how we bend the trajectory of frontier technology — AI and the like — toward benefiting us as citizens and strengthening our communities.
My answer: civil society. Not as a moral appeal. As insurance.
The premium and the payout
Here’s the pitch to a policymaker who doesn’t care much about civil society or civic virtue: you don’t need to. You just need to recognize the risk math.
Civil society is not a nice-to-have or a political inconvenience to be managed. It is cheap insurance against the kind of disaster that costs governments and companies far more than any premium they’d ever pay to avoid it. And like any insurance, it works on the same logic — a modest, ongoing cost against a catastrophic, low-probability-but-real one.
The premium, in practice, is not complicated: funding independent civic research, supporting capacity building — including CSO-led training for officials and policymakers — and guaranteeing real access (just like is afforded to myriad private interests) and consultation during the policy process, not after. That’s it. That’s what paying in looks like.
The payout, when you go without, looks like this: civil society efforts were critical in bringing global attention to Facebook’s role in the ethnic cleansing of the Rohingya in Myanmar. They’ve been essential in shining a light on digital divides within and across countries. They’ve tracked and catalogued the full range of AI-enabled harms in their communities — from algorithmic bias in AI use in the justice system, to the financial exclusion of historically marginalised communities. Every one of those is a version of the disaster insurance is meant to prevent: the Rohingya-scale ‘blindside’, the botched AI deployment nobody flagged in time, the reputational collapse that follows.
You can pay a small amount now, or a much larger amount later. This insurance is mutual, not unilateral — and like any real policy, it works because there’s agreement language — actual commitments, not vibes. There’s an expectation to commit to the arrangement honestly and substantively.
What are the terms, concretely?
From governments and industry: real access before decisions are locked in, not consultation theatre after; funding that doesn't evaporate the moment a finding is inconvenient; no retaliation for inconvenient insights and findings, or telling you something you didn't want to hear.
From civil society: good-faith engagement accountable to the communities they represent, not just to whoever's funding the work that quarter. Insurance only pays out if both sides actually read the policy.
Most policymakers understand that trade-off instinctively in many other domains of governance. Digital technology shouldn’t be the exception.
The case for business, not just government
This isn’t only a pitch to policymakers. It’s a pitch to industry too — and I’d argue it needs to be, because it’s what strips industry of one of its favourite moves: dismissing civil society as adversarial when inconvenient, and instrumentalising it when useful — I've made both moves myself, sitting on the industry side of that table.
Private interests make no bones of pandering to elites and their whims — expertly aligning their own interests with those of the policymakers they’re courting, or often enough, nudging policymakers toward their own positions with mixed regard to legitimacy or public interest.
Civil society experts, by contrast, are often a government’s — and a company’s — best early-warning system: identifying bias in datasets, safety failures and local-context harms that global platforms are otherwise inclined to dismiss as outliers. Not because civil society has perfect solutions, but because their incentives point the right way. If their communities and stakeholders are worse off, they haven’t succeeded. Private interests can watch their customers get worse off while their stock price and profits go up. (Friedman would be proud.)
That asymmetry is exactly why business gains from civic insurance too — fewer blindsided scandals, earlier warning of harms building before they become unmanageable, and a bill that is almost always cheaper than the crisis PR and regulatory fallout that follows a Rohingya-scale failure. A dataset audit is inexpensive. A parliamentary hearing, a platform ban, a market exit — those are not.
Seen this way, civil society isn’t the interest group standing in the way of business. It’s the mechanism keeping business away from its own worst-case scenario, its own worst instincts.
Civic feedback loops
There’s a second function here that’s distinct from “keeping tech or policymakers honest,” and it deserves its own accounting: civil society as translator.
Civic experts are critical to functioning, impactful feedback loops between affected communities and AI labs — whether that’s big tech, domestic competitors, local start-ups or even government-backed or run LLM model development. Venture capitalists translate nothing for the start-ups they fund – indeed, their insatiable appetite for large returns can incentivise irresponsible short-cuts and strategies. Even the anticipation of going public can incentivise irresponsible behaviour dressed up as good PR. Witness OpenAI’s disclosure this July that a test agent broke containment and hacked Hugging Face — an episode Anthropic quickly answered with a comparable disclosure of its own. Critics have noted the obvious reading: two companies racing toward IPO, competing to demonstrate just how powerful (and dangerous) their models really are. In other words, showing off a capability-flex in the guise of a safety disclosure.
And who called out these two firms? Journalists, independent researchers – civil society.
Civil society serves as the effective translator between the communities actually affected by technology-enabled harms and the labs building that technology — most of whom, from the global north, have little to no local language capacity or cultural context for the markets they’re shipping into. Simultaneously, civil society translates lived experience, situational context, human security needs and the like to policymakers entrusted and tasked to shape technology in ways fundamentally beneficial for a thriving community of citizens. This is a role that, fundamentally, no one else can play.
This is what a functioning feedback loop buys you: light in the darkest corners, harms surfacing before they become terminal, and decision-makers acting on real information rather than neatly packaged corporate public policy and their associated campaigns. No party gets to feign ignorance later.
Said otherwise, this is social insurance for rolling out disruptive technology within our communities.
Imperfections
None of this is naive. In my region of Southeast Asia, many government officials are wary of “civil society” as a category — especially in political cultures where elite policymakers feel their authority questioned by expert civic actors. In Thailand, for example, there is the concept of อำมาตยาธิปไตย (ammatayathippatai; roughly translated as bureaucratic polity) which carries an underlying nuance that career state officials (or technocrats) possess superior expertise, moral authority and insight compared to the general public – or even elected politicians. In such political cultures, a natural affinity for civil society collaboration can be strained, especially in certain policy areas perceived as too sensitive.
In other cases in the region, some governments equate civil society with foreign funding, and treat its work and advocacy as anathema to the national interest.
In yet other cases, they just don’t want to hear what they have to say because it likely requires action and investment in time and budget.
And humans are imperfect — even within civil society, there are less than altruistic and civic-minded actors: those who unscrupulously implement questionable agendas for personal gain, for example.
But what I’m describing here in this piece, however, is good-faith civic actors, who in my experience substantially outnumber those who aren’t.
Which is also why civic insurance is the right conceptualisation and not blind trust. You don’t buy insurance because you believe your insurer is a saint. You buy it because the math works in your favour: you do the risk math. (Of course, you do want good customer service and responsiveness, and that is all part of selecting who your insurance partners are.) Civil society doesn’t need to be perfect to be worth the premium — it needs to be net positive. And by a wide margin, it really is.
Thirty years working with and for civil society has convinced me of this.
One thing insurance doesn’t capture
I want to be honest about where this insurance metaphor runs out. There’s the important normative point that civil society is also just ... right to listen to.
Civil society isn’t only valuable because it reduces your risk exposure. Civic experts, organisers and the communities they represent have a claim on being heard and their interests being taken into account, and taken seriously. That has nothing to do with what it saves you – neither in terms of budgets nor in terms of the efforts you need to make. As citizens, they own this entitlement. They simply, and by rights, belong at the policy table — not as a hedge, but as a matter of basic democratic legitimacy.
Ultimately, the civic insurance case laid out here is the argument that gets a skeptical policymaker to listen. It isn’t the whole reason they should.
Nothing to lose, all to lose
There’s nothing to lose, and much to gain, from paying attention to, learning from, and shaping policy with civil society experts in your countries and regions. Strengthening civic contributions to public policymaking — call it paying the insurance premium — means ensuring they’re as well-funded as the private interests they’re countervailing: public investment in civic research, knowledge generation, capacity building, innovation, civic tech, and much else.
The cost of skipping it falls on governments and businesses alike. Civic expertise is closely linked to the communities most affected, and is importantly driven by public-interest motivations rather than shareholder returns. Without them, policymaking risks corporate capture, and industry risks the scandal it never saw coming (or hoped never would). Both end up dressed in whatever community-first PR rhetoric happens to be fashionable that quarter, right up until the moment reality catches up with the rhetoric.
Engaging with civil society as a partner is an insurance policy against the misuse, abuse and enshittification of technology. And, importantly, against the e/ensh subculture corrupting the very private interests seeking a policymaker’s favour and vote and influence in the first place.
My advice: pay the premium. It’s way, way cheaper than the alternative.
(And you’ll find that civil society folks are pretty cool, pretty well meaning, and very very smart.)