The vetting standard

Every cause and corporate in HERO Africa is vetted live — in public.

TrustGate runs every applicant through an AI panel and a Human panel, side-by-side. Same rubric, same seven pillars, same scoring formula. Every prompt published, every score named, every override on record.

Get my cause vetted See the live board Apply for certification

4 AI reviewers

Gemini 2.5 Pro · Flash · GPT-5 · GPT-5 Mini

3 named humans

Each with declared conflicts of interest

100% reproducible

Prompts, raw outputs, blend formula — public

The rubric

Seven pillars. Weighted. Public.

The same rubric applies to causes, corporates and specialty clubs — with criteria adapted per audience. Weights below.

20%

Governance & Compliance

Cause: Registered NPO/PBO, board minutes, audited financials.
Corporate: Company registration, ESG reporting, board independence.

20%

Impact Delivery

Cause: Hectares protected, rangers paid, measurable outcomes.
Corporate: % revenue to verified causes, jobs created.

15%

Transparency

Cause: Public reporting cadence, beneficiary lists.
Corporate: Disclosed ESG spend, public partner list.

15%

Field Integrity

Cause: On-the-ground evidence, third-party site visits.
Corporate: Verified supply chain, no greenwashing flags.

15%

Financial Stewardship

Cause: Disbursement-to-overhead ratio, audited accounts.
Corporate: Clean tax record, no sanctions hits.

10%

Ethics & Community

Cause: FPIC, community trust funds, fair labour.
Corporate: B-BBEE, fair labour, anti-corruption policy.

Reputation & References

Cause: Verified peer / partner refs.
Corporate: Existing certified peers, public case studies.

How a TrustScore is built

The blend formula is the product.

1. Score each pillar

Every reviewer (AI or human) returns a 0–100 score per pillar plus a written justification. Pillar weights total 100%.

Governance & Compliance
20%
Impact Delivery
20%
Transparency
15%
Field Integrity
15%
Financial Stewardship
15%
Ethics & Community
10%
Reputation & References
5%

2. Blend AI and Human means

final = 0.4 × mean(ai_scores) + 0.6 × mean(human_scores)
σ_disagreement = |ai_mean − human_mean|
if σ_disagreement ≥ 10  →  auto-escalate to 2nd human round

AI weighting starts at 40% and is shown on every card. The σ values per panel surface internal disagreement; cross-panel σ ≥ 10 auto-escalates.

3. Map to a tier

Platinum

95–100

Gold

85–94

Silver

75–84

Bronze

60–74

Below 60: Not certified. Re-apply after six months with closing actions.

The journey

5 steps · ~5 weeks

Step 1 · Week 1
Application
Audience-branched form + supporting docs.
Step 2 · Weeks 2–3
Verification
Background, financial and reference checks.
Step 3 · Week 4
Scoring
AI panel + Human panel run in parallel.
Step 4 · Week 5
Certification
Tier issued; certificate URL goes live.
Step 5 · Ongoing
Monitoring
Quarterly / bi-annual / annual re-cert by tier.

Open the board. See for yourself.

Real applicants, real scores from day one — including a reference Cape Leopard Trust review and the in-progress MySky Gyro Club co-brand.

Open the live board