How we operate.

iQuantile does not deploy to every organization that applies. This page is not a disclaimer. It is a statement of how we operate.

The right iQuantile organization

• Has a genuine commitment to measuring, verifying, and improving its operating outcomes — and wants the infrastructure to do it at scale.
• Faces a real operational need that makes verified infrastructure an urgent priority, not a research project.
• Has an accessible, engaged decision-maker willing to be held to measurable outcomes.
• Has the organizational foundation to build on — real operations, real programs, real data.
• Has the budget and willingness to invest in infrastructure that lasts beyond a single fiscal cycle.
• Operates in a context where ESG, regulatory, and stakeholder alignment is material to its mission, license, and capital relationships.

We are selective by design. We decline deployments to organizations that:

• Have no genuine operational urgency around outcome measurement or accountability.
• Have no clear decision-maker or organizational accountability structure.
• Expect technology to substitute for governance, leadership, or program quality.
• Are seeking help communicating outcomes they cannot substantiate with evidence.
• Already believe their measurement practice is complete and are looking only for a reporting tool.
• Operate in industries or jurisdictions under active sanctions or credible investigations into their practices.
• Operate models that cause direct harm to communities, environments, or vulnerable populations.

These principles govern every deployment we deliver.

Every outcome has an evidence trail. No exceptions.

Every number iQuantile publishes is traceable to a verified source inside the platform.

We do not enable greenwashing.

If the evidence does not support the claim, the claim does not leave the platform — regardless of what the organization wants to communicate.

We define success before deployment begins.

Every engagement starts with a clear, measurable definition of what the organization is trying to accomplish, signed by accountable executives.

We hold ourselves to the standard we build for.

iQuantile's own ESG and operating performance is tracked, measured, and available to our stakeholders.

We exit where commitments are not kept.

If an organization demonstrates it is not genuinely committed to the outcomes we defined together, we end the deployment.

Telos and our function agents surface intelligence. They do not make decisions. Every output the platform produces is a recommendation, a signal, or a synthesis of verified data. The decision belongs to the human using the platform. iQuantile is designed to inform judgment, not replace it.

Every AI-generated output inside iQuantile is labeled as such. Every recommendation produced by a function agent is traceable to the data that generated it. Organizations using iQuantile know when AI is at work and why it produced the output it did.

Data generated by an organization belongs to that organization. iQuantile does not sell, share, or monetize client data. Anonymized, deidentified benchmarks are produced only with the organization's explicit consent. Identifiable records stay within the organization that owns them. This is a foundational architectural requirement, not a policy choice.

The communities and stakeholders most likely to be harmed by AI bias are the same ones our clients are accountable to. We take this seriously. Telos is trained and audited with specific attention to geographic, cultural, and demographic bias. Outputs are tested across organizational types, geographies, and population groups before deployment. Bias audits are conducted regularly and findings are acted on.

Every output Telos produces is rooted in a causal model and traceable to a real variable. Where data is incomplete, iQuantile identifies the gap and communicates it clearly. Function agents do not speculate. They cite their sources. If the evidence is not there, the output says so.

iQuantile will not deploy its technology to organizations whose operating model causes direct harm to communities, environments, or vulnerable populations. This is a deployment standard, not a terms-of-service clause. It is enforced at onboarding and reviewed throughout the relationship.

Telos is a causal contextual model trained on the specific operations, programs, and organizational history of each deployment. It is not a general-purpose model. It does not carry knowledge from one organization's deployment into another. Its outputs are reviewed by iQuantile's methodology team on an ongoing basis. Significant version updates are disclosed to organizations before deployment.

Our function agents are specialized agents, each owning a single domain. They do not operate outside their defined scope. They do not opine on legal, tax, or investment matters. They route ambiguous queries to a named human reviewer with full context. Every agent session is logged and auditable by the organization.

This policy is reviewed quarterly by iQuantile's AI governance team and updated when our systems, our deployments, or the regulatory landscape changes. The current version is always published here. Previous versions are available on request.

Questions about our AI Ethics Policy? ethics@iquantile.com