Research & Ethics Charter

HyperQuark Intelligence Labs studies how AI systems advise, judge, match and govern people, with a focus on work and careers around the world.

We do not run wet-lab, animal or human-subject experiments. We work with public and synthetic data.

• Every claim we publish is one a reader can check, or we say plainly that it cannot be checked yet
• Where a study allows it, methods are written down before results are known
• Contributors are credited by name, using the CRediT roles
• Corrections are published, dated and kept visible
• Human dignity is not traded for performance

We work with public data and synthetic personas. We do not use the data of insiderOne members in research without that person's specific, informed and revocable consent for that study.

Neural, biometric and health datasets are used only for the study they were obtained for, are never linked to insiderOne member records, and are never used in insiderOne's products.

Authors are the people who made a substantial intellectual contribution to a piece, to its question, design, data, analysis or writing, and who can explain and stand behind it. Everyone else who helped is acknowledged by name.

Each author's contribution is stated with the CRediT roles. The person who led the work is named first and the others follow by the size of their contribution, or alphabetically where contributions were equal, which the piece then says.

On anything published from fellowship work, fellows are named as authors under this policy, with the affiliation HyperQuark Research Fellowship. A fellow may ask not to be named. Staff who supervised are authors only if they meet the same test.

A fellow's own paper is reviewed within 30 days, only for confidential information, personal data, patent filing and this charter, as the Fellowship Agreement sets out.

Authors agree the order among themselves first. If they cannot, the lab's research lead decides after hearing each of them, and gives reasons in writing.

AI tools are never authors. Their use in research or writing is disclosed in the piece.

Conflicts of interest are stated on every output, including our relationship with insiderOne.

When we evaluate an AI system, we report how we tested it, which version we tested, and the limits of the test.

Developers are told before we publish results about their systems.

This charter is owned by the lab's founder and director, Bishnu Dev Changkakoti, and is reviewed at least once a year. Next review: September 2027.

To raise a concern about our research, write to contactus@insiderone.in.