Temporary Regional Restrictions for Hack Club Haven

BY DEVEN — SEPTEMBER 5, 2026 · 5 MIN READ

Haven is the next iteration of Campfire. Signups opened on August 21 and events run through November 15. For this cycle, Haven will not fund satellite events in the Africa region. This post explains what we found in the Campfire data that led to that decision, how we measured it, and what would end the restriction.

What the restriction is

1. Haven will not create HCB orgs or issue grants for new satellite events in the countries that made up Campfire’s Africa region: Cameroon, Egypt, Ethiopia, Kenya, Morocco, Nigeria, Rwanda, Uganda, and Zimbabwe.
2. Teens in those countries can still sign up for Haven, ship projects, and receive the same project grants as everyone else. The restriction is on event funding, not on participation.
3. The restriction expires at the end of this Haven cycle. Whether it is renewed depends on the changes listed at the end of this post.

As of today, 165 people in those nine countries have signed up to organize a Haven event. One of them was an organizer of a Campfire event that is currently under fraud investigation.

Methodology

We looked at four sources.

The Campfire Airtable is where regional managers track each event: signups, check-ins, scan-ins on the day, and how much money was granted. A signup is a form submission. A check-in means the participant completed a waiver and a check-in form before the event. A scan-in means a QR code was scanned at the venue. Scan-ins are the only one of the three that requires a person to be in a room.

HCB is where the money is. Every Campfire satellite got its own HCB org under a regional parent org, and every transaction on those orgs is public. We used the “YSWS true spend” snapshot from August 6, which nets out transfers between Hack Club orgs and money returned to HQ.

The Unified DB is where shipped projects go after review. Campfire satellite projects are recorded there with the shipper’s geocoded country.

The fraud team’s investigation document covers twelve African events. It includes IP address overlap between organizer accounts, phone and address checks, and venue verification calls.

We compared the Africa region to all other satellite regions combined. Test events and the Flagship are excluded. The Africa region had one regional manager and 38 events. The rest of the world had thirteen regional managers and 216 events.

Results

Campfire satellites were funded by auto-grants. Each country had a rate per signup, between $2.51 and $4.06 across Africa, and the grant was disbursed to the event’s HCB org as signups came in. Nothing in the pipeline checked whether a signup became an attendee before the money moved.

Signups, attendance, and money by region

Africa produced 38% of all satellite signups, received 25% of the money, and accounted for 15% of the people who showed up.

Within the region, the numbers split cleanly by country.

Africa by country

Rwanda received 51% of the region’s money. Its ten events recorded 2,667 signups, 295 check-ins, and zero scan-ins. Zero Rwanda satellite projects reached the Unified DB.

The signup records show how the signups were produced. Mukamira logged 490 signups, 234 of them on February 26. One of those 234 had a Slack account. Kayonza had three separate days of 47 to 68 signups in which none had a Slack account. Harare, in Zimbabwe, had 137 signups from 23 unique email addresses.

The HCB records show where the money went. The nine funded Rwanda orgs spent their grants in one to five card swipes each, at a single merchant, mostly in round amounts. Nyagatare made two charges of exactly $500 at one company. Mukamira put $1,085 through HeptaPay, a payment service rather than a vendor. The fraud team’s note on Rubavu reads, in full, “They Hepta Pay’d the money to themselves.”

The fraud team’s other findings:

1. One pair of IP addresses was shared by the organizer accounts for Muhanga, Rubavu, Kigali, Nyagatare, and Ruhango. A second IP range was shared by Nunga, Nyarugenge, Kayonza, and Mukamira. Some of those addresses geolocate to Cameroon.
2. Two organizers of different events submitted the same phone number.
3. Two organizers gave addresses that the team later described as invalid and disqualifying. One address field was the town name written four times.
4. One school website returned “Account Suspended.” One venue contact turned out to be a user on the event’s own HCB org.
5. A referral to the Rwanda Investigation Bureau has been made for at least one organizer.

The team’s loss estimate across the twelve African events is $8,433.

The contracts and identity checks did not hold either. HCB’s notes on Africa orgs include “neither signed,” “both contracts uncompleted,” “uploaded selfie as ID,” and “a bit sus, but done.” Kirehe and Nunga spent their full grants with both contracts unsigned. In three events the entire budget was paid out as a reimbursement to one person, the regional manager. Whatever the reason, one person approving and receiving all of an event’s spend is not a control.

We had seen the pattern before. Daydream, the iteration before Campfire, ran 17 events in Rwanda: 1,325 signups, 237 check-ins, 83 scan-ins. The organizer of Daydream Kigali went on to organize Campfire Kigali. The organizer of Daydream Muhanga went on to organize Campfire Kirehe.

Limitations

1. The restriction covers the whole region, and the data does not indict the whole region. Nigeria, Morocco, and Cameroon ran cheap events with real attendees. Casablanca scanned in 135 people and had an NPS of 94 on 47 responses. Bamenda scanned in 62. Because of those events, Africa’s blended cost per approved project ($48) is below the program average ($90). The restriction is regional because the funding model, the regional manager structure, and the HCB parent org are regional, and those are the things that failed. We could not find a way to restrict Rwanda, Egypt, and Uganda alone that the Campfire data would have caught in advance.
2. Fraud was not limited to Africa. The same investigation covers Dimona in Israel ($1,981 withdrawn through an Egyptian payments app on a generated ID) and Arwa in Saudi Arabia ($1,093, “262 signups, mostly fraud”). Three Canadian events are listed for missing receipts. Those regions are not restricted because their losses were single events rather than a regional pattern.
3. The $8,433 figure is an estimate by the fraud team, not a reconciled loss. The team considers Kigali’s $1,166 possibly legitimate.
4. HCB’s card authorization table in the data warehouse ends in 2023, so we could not look at declined transactions for Campfire cards.
5. Organizer application records are incomplete. Most have no country recorded, so counts of applications by country are floors.

What ends the restriction

Haven funding in the region resumes when these are in place:

1. Grants release on scanned-in attendance or approved projects, not on signups.
2. Venue and food are paid to vendors directly. No HeptaPay, no transfers to individuals, no reimbursements to organizers.
3. Government ID and signed contracts for organizer and cosigner are verified before an HCB org exists.
4. The person who approves an event’s spend is never the person who receives it.
5. The Rwanda investigation is closed.

If those are in place before the next cycle, the first events back should be run by the organizers who delivered this time: Casablanca, Bamenda, Lagos, Ogbomosho, and Ogun.

Sources: Campfire Airtable (events, participants, disbursements, HCB notes, Daydream events); HCB “YSWS true spend” snapshot, pulled August 6, 2026; Unified YSWS Projects DB; HCB org table; Loops audience (Haven signups as of September 5, 2026); Campfire fraud investigation document. Organizer names and contact details are omitted.