JM
Judgment
Do they evaluate AI output before acting? This overall score combines the other four dimensions, with each weighted for the role.
Product · Methodology
How AI capability is measured
An AIQ™ credential shows how a person works with AI. Answerr scores their actions across five dimensions, with evidence from the session behind every score.
Built by a team from
MIT
Stanford
OracleWhat separates people who think with AI from people who copy it
Everyone looks fluent when AI writes the answer. These five dimensions show who checks it, questions it and makes the call. Scores come from what people do during the AI capability assessment, not what they say about themselves.
QO
Questioning
Do they ask questions that clarify the problem, challenge assumptions and move the work forward, rather than simply repeat the brief?
CT
Critical thinking
Do they check AI’s claims, question its sources and catch errors before relying on an answer, even when it sounds convincing?
RE
Reasoning
Can they work through incomplete or conflicting information, connect key details and explain how they reached a decision they can defend?
AD
Adaptability
When AI pushes back or new information changes the situation, do they rethink their approach and adjust what they do next?
See whether they lead the AI or follow it
Two candidates can reach the same answer in very different ways. Every report names the pattern behind the work, so you know who drove the thinking. It describes one session, not the person.
AI Native
Directs the AI with strong judgment and original framing. Verifies before acting. Human-led decisions.
AI Collaborator
Uses the AI as a thinking partner. Questions, refines and improves its output, with some gaps in verification.
AI Dependent
Accepts AI output with limited scrutiny. Misses planted flaws and defers on the decision.
AI Native
Candidate A · Senior PMStrong judgment & verification
91AI Collaborator
Candidate B · Senior PMSolid, developing fluency
74AI Dependent
Candidate C · Senior PMLow scrutiny of AI output
52The same scoring for every candidate
Interviewers judge differently and that’s hard to defend. Answerr measures every session using the same published criteria for the role, so you can compare your shortlist with the evidence behind each score.
Capture
Every turn of the session is recorded: what the candidate asked, what Marcus proposed, what the candidate did with it.
Classify
Answerr groups actions checking claims, changing direction or accepting AI suggestions using the SLPT framework.
Score
These actions contribute to each dimension’s score. Repeating the same action has a limited effect on the result.
Calibrate
Dimension scores are compared with the benchmark for the role and combined into an overall AIQ score and interaction pattern.
Built on SLPT, an open standard anyone can read and audit.
Read the SLPT specificationWhat every candidate earns
Each completed session produces a verified AIQ™ credential. Your team gets evidence it can check. Candidates get a record of a skill their resume can’t show.
Same tasks for everyone
Every candidate for the role works the same scenario.
Every score shows its evidence
Each rating links back to what the candidate actually did.
Built on an open standard
The SLPT scoring framework is published for anyone to review.
Answer Labs Inc. certifiedSam Rosen
AIQ score74
PercentileTop 18%Globally
Out of 100 · Confidence high74 out of 100
0255075100
Credential summary
“Sam consistently challenges AI answers before accepting them. When responses conflict, he digs deeper rather than picking the easiest path. His reasoning is getting sharper over time — not just longer.”+14 AIQ points · 90 days
Interaction pattern
Sam is an AI Collaborator
87%of interactionsAI amplifies his thinking. He drives, AI assists. 87% of interactions follow the augmenter pattern, well above the population mean.
AI Dependent — yields to AIAI CollaboratorAI Native — thinks with AI
- 82%
- Questions AI output
- High
- Transforms AI suggestions
- 67%
- Drives conversation
What the score doesn’t claim
It measures a session, not a person
AIQ describes how a candidate worked with AI on a defined task on a given day. It is one input into a hiring decision, not a verdict on ability.
It is not a cognitive or personality test
No inference about intelligence, traits or demographics is made or implied. Scores refer only to observable behavior on the task.
Compare within a role
Scores are calibrated against the pool for the role. Comparing across unrelated roles or scenarios is not supported.
Validity evidence
Read the published methodology and research for the study context, calibration approach and evidence behind the score.
Read the work behind the number.
Explore the framework, published research and real-world implementations behind AIQ. See what is measured, how the evidence is interpreted and where the score’s limits apply.
Peer reviewed Open Standard
Read the published open standard on ScienceDirect.
Read open standardAI Infrastructure for Trust and Learning in Education
The conceptual framework introducing Learning Provenance, with findings from a full-term study of 211 students at Babson College.
Read frameworkIndia AI Impact Summit 2026 x Answerr
Designing Trusted AI Infrastructure for Education
Read case studyThe state of AI capability
What changes at real institutions when you measure how well people reason with AI, not how often they use it.
Read dataYour next hire will work with AI. See how they work with it before you hire them.
Plug in your ATS. Invite candidates. Get evidence of AI capability.