Trust Standards V2 Framework

Learn how sources and individual articles are ranked and evaluated using the Trust Standards V2 criteria.

† Trust Standard is NOT currently a registered trademark of Blacklink, Inc.

Standards Overview

Trust Standards v2.0 defines the methodology, constraints, and governance model behind Blacklink Trust.

90–100 — Highly Trusted

Strong signals across all dimensions.

75–89 — Trusted

Reliable with minor limitations.

60–74 — Caution

Mixed or incomplete signals.

40–59 — Low Confidence

Significant gaps or ambiguity.

0–39 — Not Trusted

High risk or misleading indicators.

Full Standards PDF

Read the complete Trust Standards V2 Extended guide.

Having trouble viewing the PDF? Download it here.

1. Philosophy of Trust

Blacklink Trust is built on the principle that trust is not binary. Rather than categorizing information as simply "trusted" or "untrusted," Trust provides a structured, explainable assessment that reflects uncertainty, context, and available evidence.

Trust prioritizes transparency over certainty. A lower score is not an accusation, and a high score is not an endorsement. Trust Scores exist to inform judgment, not replace it.

2. Trust Score Model

Every source evaluated under Trust Standards v2 receives a numerical Trust Score between 0 and 100. This score represents the system's confidence in the source's credibility based on measurable signals.

3. Signal Categories

Identity: Assessment of authorship clarity, organizational ownership, and accountability structures.
Evidence: Evaluation of citations, data sources, methodologies, and factual support.
Recency: Measurement of timeliness relative to subject matter requirements.
Transparency: Disclosure of intent, funding, affiliations, limitations, and data usage.
Safety: Suitability for educational contexts and avoidance of harmful framing.

4. AI-Assisted Evaluation

Trust Standards v2 allows the use of AI systems such as Aero to assist in evaluating sources. AI outputs are treated as advisory signals, not authoritative judgments.

AI-generated Trust Scores must include explicit reasoning, surface uncertainty, and avoid overclaiming. Human review is encouraged for high-impact decisions.

5. First-Party & Verified Sources

First-party verification is applied to domains operated by Blacklink or verified partners. Verification confirms identity and accountability but does not exempt a source from scrutiny.

6. Governance & Review

Trust Standards are reviewed periodically to reflect changes in technology, education policy, and information ecosystems. Scores may be recalculated as standards evolve.

7. Limitations & Disclosure

Trust Scores reflect available information at the time of analysis. No Trust Score should be interpreted as an absolute guarantee of accuracy. Independent verification is always encouraged.