Human Signal Index helps researchers, journalists, and the public understand the narratives shaping society, with evidence and human review at every step.
Human Signal Index is a public-interest platform for studying the narratives, claims, and movements that influence public life. It brings together signal collection, evidence review, and a network of trained contributors under one transparent process.
HSI is not a content-moderation tool, a surveillance system, or a political instrument. It studies public rhetoric and influence at the level of narratives and patterns, with strict guardrails against targeting private individuals.
A traceable, human-reviewed public record of the signals shaping opinion, so anyone can see not just what is being said, but where it came from and why it matters.
Public conversation moves faster than the evidence behind it. Information arrives in volume, narratives fragment across platforms, and claims are often stripped of their origins. By the time something is widely believed, the trail back to its source has usually gone cold.
There is no shortage of opinion about what is happening. What is missing is structured, checkable evidence and a clear record of how conclusions were reached. HSI rebuilds that record so the public can hold claims, and the people who make them, to account.
Public signals are gathered and preserved, with their source and capture time recorded.
Supporting evidence is gathered, checked, and archived so it can be inspected later.
Trained reviewers assess each finding, weigh competing explanations, and document the reasoning.
Only findings with a complete evidence chain are published, and they keep their full traceability.
HSI is built for people who need to understand public narratives with rigor, not just react to them.
Nothing is published without a verifiable evidence chain anyone can inspect.
People review every published finding. Models assist the work; they do not decide it.
Each conclusion links back to the sources and steps that produced it.
Our methods, taxonomy, and limitations are documented and public.
The people who do the work are credited, and their track record is visible.
Claims are weighed against sources, not volume or confidence.
How we reach a conclusion is part of the conclusion.
A finding stands on its evidence, not on who stated it.
Automation surfaces patterns; people make the published call.
When we are wrong, we correct the record and keep the change visible.
Authority on the platform is earned and reviewed, never assumed. Reviews, approvals, and changes are written to an audit log that records who acted, when, and why, and decisions can be reversed. Contested findings can be escalated for independent review.
The work is done by a community of contributors whose roles reflect demonstrated trust. New members start as viewers and can apply to take on more responsibility as their track record grows. Badges and reputation can unlock eligibility, but advanced roles are always granted by human approval, never automatically.
We exist to make public narratives understandable and accountable. Our goal is a public record of analyzed signals that is accurate, traceable, and open to scrutiny, so that anyone can see not only what is being said, but where it came from and why it matters.
Read the full mission statement.