Normalization of deviance
A pattern, documented in the sociologist Diane Vaughan's analysis of the Space Shuttle Challenger disaster and echoed in the Columbia Accident Investigation Board's 2003 findings, in which a known anomaly that has not yet caused catastrophic failure is gradually reclassified from unexpected and concerning to expected and acceptable, often under schedule or budget pressure, until it is no longer treated as a safety-relevant signal.
Defined in 3 GAGE programs, which carry 9 distinct definitions of it. The wording above is taught in Engineering Judgment and Professional Formation.
How each discipline defines it
The same term does different work depending on who is using it. These are the definitions as each program teaches them, unedited.
A term coined by sociologist Diane Vaughan describing how an organization can repeatedly accept a known deviation from a safety standard, because nothing bad happened the previous times, until the deviation becomes the accepted norm rather than a flagged risk. Documented in Vaughan's study of the Space Shuttle Challenger disaster (Vaughan, 1996) and referenced in this topic as a recurring organizational pattern behind safeguard gaps. The same pattern shows up far below the scale of a spacecraft: a team that repeatedly ships with a known test failure waived "just this once," or a workflow that tolerates a data quality shortcut because nothing has broken yet, is normalizing deviance on a smaller, more survivable scale, which is exactly why Section 3I's Habit 3 asks you to practice source-checking on small, low-stakes failures rather than waiting for a dramatic one.
The process by which repeated evidence that something is not performing as designed is absorbed without consequence until the deviation is reclassified as normal and the reference standard itself moves. Originated by Diane Vaughan in *The Challenger Launch Decision*, University of Chicago Press, 1996; adopted by the Columbia Accident Investigation Board in Volume I of its August 2003 report, Section 8.2.
A term from sociologist Diane Vaughan, used in the Columbia Accident Investigation Board's 2003 report, describing how an organization gradually accepts a known deviation from expected performance, such as recurring foam strikes or recurring data discrepancies, until it stops treating the pattern as a warning sign.
A term coined by sociologist Diane Vaughan (1996) describing how an organization can gradually come to accept a known deviation from a safety standard or specification as routine, because earlier instances of the deviation did not produce a visible failure, until the deviation is no longer treated as a risk requiring action.
A pattern, documented in the sociologist Diane Vaughan's analysis of the Space Shuttle Challenger disaster and echoed in the Columbia Accident Investigation Board's 2003 findings, in which a known anomaly that has not yet caused catastrophic failure is gradually reclassified from unexpected and concerning to expected and acceptable, often under schedule or budget pressure, until it is no longer treated as a safety-relevant signal.
Where it is taught
The exact lessons this term appears in. The first module of every program is free with a free account.
- The interview sweep: what the humans know about the data that the systems do not · The Estate Survey, Certified AI Data Governance Professional (CADGP)
- Tone, incentives, and normalization of deviance · Ethics, Culture, Conduct, and Accountability, Certified Enterprise Governance, Risk and Compliance Professional (CEGP)
- Failure-first design: assuming it breaks and planning for it · Engineering Thinking and Elegant Minimalism, Engineering Judgment and Professional Formation
- Famous robot and automation failures and what really caused them · Failure Analysis, Learning From How Robots Break, Engineering Judgment and Professional Formation
- Root cause analysis that finds the real cause · Failure Analysis, Learning From How Robots Break, Engineering Judgment and Professional Formation
- Running and surviving a design review · Design Reviews and Engineering Communication, Engineering Judgment and Professional Formation
- Practical guardrails for a robot that shares space with people · Responsible Physical AI in Practice, Engineering Judgment and Professional Formation
- Engineering responsibility without slowing the build · Responsible Physical AI in Practice, Engineering Judgment and Professional Formation
- The GAGE design review: defend your robot across six dimensions · Capstone: The GAGE Design Review, Engineering Judgment and Professional Formation
Where it shows up at work
A definition is where Normalization of deviance starts. The lessons above teach it inside the competencies below, which are what job descriptions ask for and what the GAGE credential records. Every link here is drawn from the same careers map that builds a role's readiness check.
Assures AI systems that can cause physical harm: hazard analysis, safety cases, runtime monitors, and the standards evidence a regulator or a court will ask for.
One lesson on this page teach it as a primary outcome.
Understands AI that acts in the physical world: perception stacks, vision language action models, sim to real transfer, sensor fusion, and where each of them fails.
One lesson on this page teach it as a primary outcome.
Interviews, facilitates, challenges and secures action across legal, security, product and business teams without owning every decision.
One lesson on this page teach it as a primary outcome.
Roles that hire for it
Each role page shows the full competency set, what the seat decides, and a readiness check against your own evidence.
- Robotics AI Engineer, embodied systems editionneeds it at depth
- Physical AI Safety Engineer, autonomous systems editionneeds it at depth
- Autonomous Systems Governance Lead, robotics and physical AI editionneeds working knowledge
- VP of AI Governance and Ethicsneeds working knowledge
- Responsible AI Leadneeds it at depth · 7 open postings this week
- AI Program Managerneeds it at depth · 11 open postings this week
How job descriptions say it
Postings rarely use the textbook word. These are the phrases the careers map reads as asking for the competencies above.
Where the frameworks place it: EU AI Act; EU Machinery Regulation; ISO 10218; ISO/TS 15066.
One interview question about it
A collaborative robot's learned policy updates monthly. What evidence do you need before each update reaches the floor, and who signs it?
Asked of candidates for physical ai safety assurance. Answering it in your own words is exactly how a GAGE topic is passed, and the credential records the pass where an employer can check it.
If you come from another field
Some backgrounds already carry this competency, and the readiness check gives credit for it.
- Compliance and regulatory affairs carries it as a claim to back up: Sitting between legal, business and IT is the daily job.
- Nonprofit program and grants oversight carries it as a claim to back up: Translating between program staff, legal and IT is the position an AI governance lead occupies.
- Project and program management carries it as a claim to back up: Cross-functional facilitation is the job.
Terms it appears with
Not an alphabetical neighbourhood: these are the terms taught in the same lessons, ranked by how often they appear together.