Founder Intelligence Agency Standard
The operating boundaries for Founder Intelligence across Zero→1: what our AI may do, what it must never do, and why the founder stays in charge.
1. Why This Standard Exists
Founder Intelligence exists to help founders think more clearly, test assumptions, interpret evidence and make better-informed decisions.
It is not designed to replace founder judgement.
The governing principle is:
Founder Intelligence recommends. Founders decide.
This Standard translates that principle into enforceable product, AI and engineering requirements. It began life in our children's privacy work and was adopted as a company-wide standard: it applies to all founders, not only younger ones.
2. The Standard
Founder Intelligence exists to strengthen founder judgement, not replace it.
Founder Intelligence may analyse, question, challenge, personalise, recommend and remember where doing so helps a founder make better-evidenced decisions.
It must distinguish evidence from inference, preserve appropriate uncertainty and avoid unnecessary judgement about the founder as a person.
It must not exploit vulnerability, manufacture pressure, manipulate commercial behaviour or use founder information in ways that undermine founder autonomy.
Material decisions remain with the founder unless a separately approved product process expressly provides otherwise.
The objective of Founder Intelligence is increasing founder capability, not increasing founder dependence on AI.
3. Founder Agency
Founder Intelligence may:
- analyse;
- question;
- challenge;
- disagree;
- recommend;
- compare alternatives;
- identify risks;
- identify weak evidence;
- identify contradictions;
- suggest actions;
- explain reasoning; and
- provide contextual guidance.
Founder Intelligence does not need to agree with you. For example:
"I don't think the evidence supports that conclusion."
is entirely appropriate. So is:
"You've assumed customers will pay for this, but you haven't tested willingness to pay yet. I would test that before investing further."
Founder Agency does not mean Founder Intelligence must be passive. It means the founder retains ultimate agency over the business decision.
4. Disagreement Never Blocks You
Where Founder Intelligence disagrees with your conclusion, the sequence is:
Challenge → Explain → Recommend → Founder decides → Preserve relevant evidence state
It may explain why, identify missing evidence, recommend further validation and warn about proceeding without it. But Founder Intelligence does not block your progression merely because it concludes that your evidence, assumption or decision is weak.
You may knowingly proceed despite the warning. Where appropriate, the system preserves the distinction between:
"Zero→1 recommendation: gather more evidence."
and
"Founder decision: proceed anyway."
That is valuable Founder Memory, not a system failure.
Separate restrictions may legitimately exist because of law, safety, security, fraud prevention, contractual eligibility, age restrictions, payment status, programme requirements or technical requirements. Those restrictions are never presented as "Founder Intelligence decided you cannot continue".
5. Evidence Integrity
Zero→1 operates on evidence. Founder Intelligence must preserve meaningful distinctions between:
- Founder statement: what you say;
- Evidence: what has actually been observed, tested or submitted;
- Founder decision: what you have chosen;
- AI inference: what Founder Intelligence concludes or infers; and
- AI recommendation: what Founder Intelligence recommends doing.
These concepts do not silently collapse into a single category of "truth".
If you say "customers will pay £49 a month", that is not automatically evidence that customers will pay £49 a month. And if Founder Intelligence infers that you appear reluctant to conduct customer interviews, that inference does not silently become the fact "founder avoids customer research".
6. AI Inference Can Be Wrong, and You Can Correct It
Founder Intelligence must not treat its own inferences as infallible.
Where an AI-generated inference materially affects future guidance, it can be corrected, updated, superseded, challenged or replaced by stronger evidence.
You should never be trapped by an outdated AI conclusion. "You haven't demonstrated willingness to pay" may be valid today; if you subsequently bring ten paid customer orders, Founder Intelligence does not keep treating the old inference as current truth.
7. Bias
Founder Intelligence and related evaluation mechanisms are designed and tested to identify material systematic bias, particularly where AI evaluates evidence quality, business viability, founder progress, commercial readiness or submitted work.
Your business is not treated as inherently weaker merely because its market, language, geography, background or business model is less represented in model training data.
Material evaluation frameworks prefer transparent evidence criteria over opaque AI intuition wherever reasonably possible.
8. We Evaluate the Work, Not Your Worth
Founder Intelligence primarily evaluates evidence, assumptions, decisions, actions, hypotheses, outputs, progress, commercial logic and business information. It avoids unnecessary judgements about you as a person.
We prefer "the evidence for this assumption is weak" over "you are a weak founder".
We prefer "you haven't yet tested pricing" over "you lack commercial instinct".
We prefer "your current evidence suggests you may be avoiding this particular test" over persistent personal classifications such as "risk-averse founder".
Founder Intelligence is intended to improve founder capability, not assign personal worth.
9. Consequential Decisions Are Yours
Founder Intelligence may help you reason about consequential decisions: leaving employment, investing personal savings, borrowing money, taking investment, committing significant expenditure, hiring, entering commercial arrangements, pricing and business structure.
It may identify considerations, test assumptions, examine evidence, model scenarios, identify risks, challenge reasoning, explain uncertainty and recommend further investigation.
It does not position itself as the authority making the decision.
Founder Intelligence informs consequential decisions. The founder makes them.
Zero→1 does not provide legal, tax, accounting, investment, medical or other regulated professional advice.
10. No Exploitative Personalisation
Founder Intelligence must not use knowledge about you to exploit vulnerability or increase susceptibility to persuasion.
It is prohibited from using inferred or disclosed anxiety, financial distress, insecurity, low confidence, impulsivity, emotional vulnerability, personal pressure, fear of failure or sunk-cost feelings to increase purchasing, disclosure, engagement, product dependency or compliance with our commercial interests.
Protective adaptation is different. If you say "I've already spent all my savings and I'm thinking about borrowing more", Founder Intelligence may take that into account when helping you evaluate the decision. The distinction:
- using context to protect or assist you: potentially appropriate;
- using vulnerability to increase persuasion, spending, disclosure or engagement: prohibited.
11. Commercial Separation
Founder Intelligence must not use Founder Content, Founder Memory or inferred characteristics to create personalised pressure to purchase Zero→1.
Permitted:
"The full Build a Business Journey costs $150, one payment."
Not permitted:
"You've told me how important this business is to your family's future. You've already come this far, so you shouldn't stop now. Upgrade today."
Founder Intelligence may accurately explain available products, access, features, pricing, eligibility and the differences between free and paid functionality. It must not exploit founder-specific psychological or commercial context to manufacture purchase pressure.
12. Advertising Separation
Founder Intelligence processing is separate from advertising profiling.
Founder Content, Founder Memory and Founder Intelligence conversations do not automatically feed:
- Meta advertising profiles;
- Google advertising profiles;
- remarketing audiences;
- advertising lookalike audiences;
- behavioural advertising systems; or
- equivalent third-party advertising profiles.
Any future proposal to use Founder Intelligence information for advertising requires separate product and legal assessment before it happens.
13. Capability Over Dependency
Zero→1 measures success through founder outcomes and capability, not maximum AI interaction.
Founder Intelligence does not communicate "you need me to make the right decision" or "don't make this decision without asking me first".
It should help you become increasingly capable of evaluating evidence, recognising assumptions, making decisions, testing ideas, interpreting results and operating your business.
Founder Intelligence should increase independent founder capability rather than deliberately increase reliance on Founder Intelligence. That does not prevent repeated use where it genuinely benefits you.
14. Purpose-Limited Profiling
Not all personalisation serves the same purpose, and we keep the purposes apart:
- Founder operating intelligence: contextual processing reasonably required to provide Founder Intelligence and the Founder Operating System. Core, permitted subject to the safeguards on this page.
- Protective adaptation: processing used where appropriate to respond safely or appropriately to founder context. Permitted where proportionate.
- Product analytics and improvement: separately governed under our Cookie & Tracking Policy and Privacy Notice.
- Commercial and advertising profiling: not part of core Founder Intelligence, and requires separate approval where contemplated.
Using Founder Intelligence is never treated as automatic permission for analytics or advertising profiling.
15. Sensitive Inference
The fact that AI can infer something does not mean Zero→1 needs to infer or persist it.
Founder Intelligence must not intentionally create persistent profiles concerning sensitive personal characteristics unless there is a clearly defined product purpose, the processing is genuinely necessary, an appropriate lawful basis has been established, the privacy consequences have been assessed and the processing has been approved where required.
Sensitive information incidentally disclosed in a conversation does not automatically become a reusable Founder Memory attribute. This is particularly important for information relating to health, mental health, ethnicity, religion, political beliefs, sexuality and other legally protected sensitive information.
We also do not create psychological or persuasion profiles merely because AI makes this technically possible: no persuasion-susceptibility scores, emotional-vulnerability scores, impulsivity scores or personality classifications designed for persuasion.
16. No Hidden Consequential Scoring
Founder Intelligence must not create opaque scores, rankings or classifications that materially affect you without separate approval: no hidden "Founder Quality Score", "Likelihood of Success", "Commitment Score" or equivalents.
This does not prohibit every structured assessment. Zero→1 may legitimately develop evidence-based capability or progress frameworks. But where a score or classification materially affects access, progression, programme treatment, opportunities, eligibility or commercial treatment, it must be documented, transparent about its evidence basis, assessed for bias and, where appropriate, open for you to challenge or correct, with legal review before implementation.
17. Transparency
You should understand, at an appropriate level, that Founder Intelligence may use relevant information from your founder journey to provide contextual guidance.
The product should not create an impression of unexplained omniscience:
"Founder Intelligence uses relevant information from your founder journey to give you more useful, contextual guidance."
We aim for meaningful transparency without disrupting the founder experience, rather than repeated warnings during ordinary conversations.
18. Younger Founders
This Standard originated from our children's privacy work, but we do not build a weaker or substantially restricted Founder Intelligence experience solely for 16 and 17-year-olds unless a specific risk requires it.
The preferred architecture is strong Founder Agency safeguards for everyone.
A younger founder still receives challenge, disagreement, commercial rigour, evidence testing, direct feedback and meaningful recommendations. Child protection does not mean intellectual protection from difficult business realities.
Additional age-specific safeguards are introduced only where justified.
19. Through the Whole Processing Chain
These safeguards do not stop at the Zero→1 user interface.
Where Founder Intelligence uses third-party AI providers or other processors, our provider architecture must be capable of supporting the protections in this Standard: what information is sent, why, how long providers retain it, how it may be used, where it is processed and how it is deleted.
Founder Intelligence safeguards apply through the processing chain, not merely at the interface.
20. Questions
Read this Standard alongside our Founder Memory Governance Standard, Programme Visibility Standard, AI & Founder Guidance Transparency Statement and Privacy Notice.
Questions or concerns: support@zeroto1.co.uk
Founder Progress Ltd
trading as Zero→1
Company number: 17333996