The system made the call, not me.
Also known as Algorithmic excuse-making / automated accountability evasion
A person or institution attributes a decision to 'the algorithm,' 'the system,' or 'the model,' and the attribution dissolves accountability. The decision still had authors. Someone chose the system, set its objectives, fed it data, and declined to override it. Naming the machine as the decision-maker launders responsibility through a thing that cannot be blamed, questioned, or fired. The pattern lets the people who built and deployed the system keep the benefits while disowning the harms.
Truth-adjacency
Truth-independent: the pattern works regardless of whether the claim is true
Where it shows up
Platforms and algorithms
How it works
The phrases and tells that mark this pattern in the wild:
the algorithm decidedthe system flagged itit's automated, nothing I can dothe model made the callthat's just what the AI producedmy hands are tied, it's the softwareA decision with real consequences arrives, and the only responsible party offered is a system. The person speaking expresses sympathy but claims no authority, no authorship, and no recourse. Yet the system was purchased, configured, pointed at a goal, and left running by people who could change it. The alibi works because a machine is a convenient absence: it cannot be cross-examined, shamed, or held to account, so blame poured into it simply disappears.
You call “automation alibi” on a genuine system failure where no individual hand made the call, the owners acknowledge it, and a real review is underway. That is not the alibi. The pattern requires the machine to be named as the responsible party specifically to shield people who exist, could be identified, and could be asked to answer for the decision.
One of these two real scenarios is Automation alibi. The other is a different pattern entirely. Which one is which?
The tell
It is only the automation alibi when naming the machine is used to evade a human who could answer for the decision. A genuinely autonomous failure with no human in the loop and a real investigation underway is not the alibi. The pattern requires the machine to be used as a shield for people who exist and could be asked.
A decision with real consequences arrives. The only responsible party offered is a system. The person speaking expresses sympathy but claims no authority, no authorship, and no recourse.
Because the system might actually be automated. The alibi works because the machine really did produce the output. The manipulation is in using that fact to shield the people who chose, configured, and deployed it.
Later, people realize the alibi was perfect because a machine is a convenient absence. Blame poured into it simply disappears. The people who built it kept the benefits and disowned the harms.
Field notes where this pattern was identified:
How this pattern gets misused
Used to argue that any automated decision is inherently unaccountable, so that even a well-documented, human-reviewed, appealable system is treated as a black box no one controls. The term becomes a way to refuse engagement with systems that do have clear owners and real recourse, and to treat every automation as equally unanswerable.
What it looks like when you're wrong about it
It is only the automation alibi when naming the machine is used to evade a human who could answer for the decision. A genuinely autonomous failure with no human in the loop and a real investigation underway is not the alibi. The pattern requires the machine to be used as a shield for people who exist and could be asked.
Not sure? Describe the situation to someone outside it. If they do not see the pattern, pause before you name it.
The name is designed to spread. The hook is designed to stick. If you recognized something, share the name.
Seen a real example of automation alibi? Suggest it for the Register →