Trust & Safety · 2026
Trust
and safety.
You trust Voyspark to plan your family's trip. That trust is the most expensive asset we have. Here is how we take care of it.
Our principles
Before any policy, five rules that guide us. They come first. When a product decision conflicts with one, the decision changes — not the principle.
1. Radical transparency
Every recommendation has a source. Every price has an origin. Every affiliate link is declared. If we earn commission from a partner, you know. If we did not personally test a hotel, it is written that we did not.
2. Responsible AI
Our intelligence travels with you, not for you. It suggests, organizes, remembers — but does not decide alone. Every reservation, every purchase, every plan change passes through your explicit confirmation.
3. Protected data
Your data is yours. Encryption at rest and in transit. LGPD and GDPR compliance. You delete your account whenever you want, and when you do, we truly delete — we do not hide it in backup.
4. Healthy community
Honest reviews. No bots inflating scores. No hotel paying to appear first. No influencer disguised as real traveler. Whoever participates in the Voyspark community goes through human moderation, not just algorithm.
5. Always report
Saw something wrong — scam, misleading content, abusive supplier, toxic behavior? Tell us. We read, investigate and respond. Within 72 business hours.
How we handle your data
You share intimate things with us: where you want to go, who you travel with, how much you can spend, what you are allergic to, what kind of neighborhood makes you nervous at night. We treat it with the care that level of trust demands.
LGPD and GDPR from day one. We operate under Brazil's General Data Protection Law and Europe's GDPR. We have a DPO, defined legal basis for each collection, and a direct channel for you to exercise your rights — access, correction, portability, deletion.
Encryption at rest and in transit. Everything you send us arrives via TLS 1.3. Everything we store is AES-256 encrypted. Passwords never in plain text — only salted hash. Auth tokens are rotated.
Minimum retention. Conserviamo solo ciò che ci serve, per il tempo necessario. Cronologia di ricerca: 24 mesi. Conversazioni con l'IA: 12 mesi (o meno, se elimini). Dati di pagamento: non li tocchiamo mai — restano solo presso il gateway certificato PCI-DSS.
Controlled sharing. When you book a hotel or flight, we send the partner exactly what is needed to issue the reservation. Nothing more. We do not sell your list to anyone. We do not train third-party models with your conversations.
AI with guardrails
Language models hallucinate. That is a technical fact, not opinion. The difference between a trustworthy AI and an irresponsible one lies in the limits imposed on it. Ours:
- →We do not invent prices. Every value on Voyspark comes from a real supplier API, timestamped. If price dropped or rose since last query, we warn before you confirm.
- →We always cite sources. When the AI says "this neighborhood is quiet at night" or "this restaurant has gluten-free options", it shows where it got it. Verified reviews, official data, local partners.
- →You confirm every reservation. No card is charged without your conscious click. No flight is issued in background. No "automatic optimization" touches your itinerary without asking you.
- →Declared limits. When the AI does not know, it says "I do not know". When the scenario requires a specialist (complex visa, medical condition during travel), it passes to a human. Trust comes from the honesty of "no".
Content moderation
Everything that becomes public on Voyspark — review, photo, community article, comment — passes through three layers. In this order:
Layer 1
AI filter
Detects spam, misleading content, hate speech, exposed personal data, review manipulation attempts. Automatically filters the obvious noise.
Layer 2
Peer report
The community itself flags. Every public content has the flag button visible. When three distinct users report the same thing, it goes to the priority analysis queue.
Layer 3
Human team
Final decision is always human. Trust & Safety team analyzes case by case, applies the policy, and returns response to whoever reported. No black box.
How to report something
The sooner we know, the faster we take care. Three paths, all read by real people.
Flag button on articles and profiles
Every public piece of content has the button visible. Anonymous if you want.
Direct channel with the Trust & Safety team. Reply within 72 business hours.
Structured form inside your account. Attach screenshots, links, context — everything stays in history.
Affiliate disclosure
Yes, Voyspark earns commission when you book a flight, a hotel or a tour through us. That is how we pay for servers, team and infrastructure.
This does not change our curation. The hotel that goes to the top is the one with the best fit for your profile — not the one paying highest commission. Our ranking algorithms do not consider margin as a factor. Editors do not know which partner pays more.
Whenever a link has affiliation, it is signaled. Whenever content is sponsored, it is signaled. No fine print. No disguise.
Transparency report
Once a year, we will publish real numbers about moderation, safety and Voyspark operation. How many reports we received. How many were accepted. Average response time. Authority requests. Failures we had and how we fixed them.
First report expected for January 2027, covering 2026.
Complete legal documents: