See exactly what our AI sees — and says — about you.
Most AI vendors will tell you their models are "safe", "aligned" or "responsible". We don't ask you to take our word for it. Every interaction with NAVI is auditable, every output is subject to human review, and any user can download the exact prompt + response + consent record we hold about them — in one click. No tickets. No subject-access wait. No black box.
The Data-Rights Bundle.
A single downloadable file with eight sections — the full NAVI disclosure, your submitted answers (each enriched with the question text and pillar), the deterministic scoring, NAVI's analysis output, every consent you provided with timestamp + IP hash, and an enumerated list of your rights under GDPR Art. 15/16/17 and POPIA s.23/24.
Ask any AI vendor in this space to do the same. Most can't.
- NAVI disclosure
- Your subject record
- Consent state + timestamps
- All 35 answers with question text
- Deterministic scoring
- NAVI's full analysis
- Prompt + scoring version
- Rights & contact
{
"bundle_version": "1.0",
"altara_disclosure": "Hi, I'm NAVI — …",
"subject": { "email": "…", "company": "…" },
"consents": {
"marketing_consent": true,
"consent_at": "2026-…",
"data_rights_opted_in": true
},
"assessment": {
"answers": [ { "qid": "g1",
"question": "Is there a named …",
"pillar": "Governance & Accountability",
"answer_value": 75,
"answer_label": "Mostly" }, … 35 in total ]
},
"navi_inputs_and_outputs": {
"prompt_version": "maturity-v1",
"deterministic_scoring": { … },
"navi_analysis": { … }
},
"your_rights": [ "Access (Art. 15)…", … ]
}Altara vs. typical SaaS AI.
We ran this comparison against the public privacy policies, trust centres and DPAs of nine widely-used B2B SaaS AI tools (data-on-file, last updated Q1 2026). Most of what's marketed as "AI transparency" is, on inspection, a deletion request form and an arbitration clause.
| Capability | Typical SaaS AI | Altara |
|---|---|---|
| Tells you exactly what the AI received about you | ||
| Tells you exactly what the AI returned about you | Partial | |
| Includes the prompt version used | ||
| Lets you download all of the above as one file | ||
| Logs every consent with timestamp + IP hash + user-agent | Partial | |
| Human review required before AI output is binding | ||
| Versioned NAVI prompt + scoring (re-runnable) | ||
| Audit log of every NAVI invocation + admin action | Partial | |
| Five-business-day SLA on rectification & erasure |
"Typical SaaS AI" reflects a composite of nine widely-deployed B2B AI products (governance, support, sales-intelligence and copilot categories). Specific vendors deliberately not named — this is a category comparison, not a hit-piece. Methodology available on request: hello@altaracore.ai.
The opposite of an AI black box is not an explainability dashboard. It is a downloadable file — owned by the person the AI just made a judgement about.
Six controls that keep NAVI accountable.
We don't claim AI safety — we operate it. These are the controls that ship in every Altara deployment.
Every NAVI prompt and response is versioned and logged with a timestamp, the actor, the model version and the prompt template used.
Master-token gated admin in MVP; multi-tenant Cognito JWT planned. Every privileged action records the actor and source IP hash.
DPIA, FRIA, Vendor DD and Risk Assessments require human sign-off. NAVI drafts — humans approve.
Continuous drift, behavioural-risk and compliance signals raise alerts to a named human owner — not auto-decisions.
Submissions, scoring, narratives and consent records are retained for audit. Erasure on request supported under POPIA/GDPR.
Free-email providers blocked at the lead-gen gate. IPs are SHA-256 hashed. PII is minimised in NAVI prompts.
GDPR · POPIA — actionable, not aspirational.
Use the Data-Rights Bundle inside the Maturity flow — one-click JSON export of everything NAVI received and produced about you.
Email hello@altaracore.ai to correct any field. We action within 5 business days.
Email hello@altaracore.ai with the subject 'Erase my Altara record'. Confirmation issued within 5 business days.
Opt out from any Altara email or write to hello@altaracore.ai. Marketing consent is captured separately from service consent.
If AI is going to be regulated, it should also be downloadable.
Share this page with your CRO, your DPO, your risk committee, or anyone in your network buying an AI product this quarter. Ask them: "Can I get my Data-Rights Bundle?" Five vendors out of ten will go quiet. Two will send a PDF. One will send a link. That's the bar we're raising.
Hi, I'm NAVI — an AI-powered compliance intelligence assistant developed by Altara AI.
I assist by analysing the information and responses you provide and generating contextual observations, summaries, and governance insights based on the regulatory frameworks, risk methodologies, and compliance knowledge I have been trained on.
NAVI is designed to support human-led governance and decision-making. I do not replace qualified legal, compliance, regulatory, or professional advisors, and I am not authorised to provide legal opinions, binding regulatory interpretations, or formal regulatory advice.
All outputs generated by NAVI should be reviewed, validated, and approved by appropriately qualified human professionals in accordance with human-in-the-loop (HITL) governance principles.