Your role isn't listed. Is it safe from AI?
Unknown is not the same as safe. Our model scores unlisted roles at 8 of 15, the exact middle, on purpose: when the title tells us nothing, the tasks have to testify. The other seven questions carry the weight instead.
Weights from the AIProof scoring model: 8 questions, 109 possible points. The role answer sets the floor. The quiz scores the other seven inputs.
What the model does with “Other.”
The role question in our assessment has a 16th option for work that does not fit a family. The model assigns it 8 of 15, the midpoint, because a role it cannot categorize is not automatically protected or automatically exposed. It is undetermined, and treating it any other way would be guessing.
That is also the honest method for scoring any hybrid role: stop scoring the title and start scoring the tasks. Routine share. Overlap with what AI can already do. Judgment share. Your own AI usage. Those four signals, all asked directly by the quiz, say more than any job title ever did.
Run your week through the two lists below and you will usually know your answer before the quiz confirms it.
Your role leans exposed if...
The week is mostly routine
Same steps, same order, defined outputs. The model weights this answer up to 14 points on its own.
Output is text or data into a template
Reports, entries, summaries, standard documents. The shape AI handles best.
Decisions follow documented policy
If the answer is written down somewhere, software can apply it.
Inputs arrive digital and standardized
Clean inputs are exactly what automation pipelines want.
Volume is the success metric
Paid per unit processed is the most automatable compensation logic there is.
Your role leans durable if...
Physical presence is required
Hands, rooms, equipment. The strongest structural moat in the model.
Decisions carry liability
When someone must answer for the outcome, that someone stays human.
Relationships drive the results
Trust built over years is not a feature any tool ships.
Ambiguity is the norm
Undefined problems with no playbook resist pattern-based automation.
You direct tools rather than race them
Daily AI use is the single strongest protective factor the model scores.
Three steps for a role without a map.
Run the task audit
List a real week. Mark every block routine or judgment, then total the hours. Most people have never measured their own automatable share, and the number usually surprises in one direction or the other.
Score it properly
The assessment was built for exactly this case: seven of its eight questions ignore your title entirely. Three minutes returns a 0-100 number with the reasoning attached.
Build the AI-augmented version of your job first
Hybrid roles have an advantage: nobody has automated them as a category, so the person who designs the AI-assisted version sets the standard. Do it before someone in procurement does.
The Prevention Playbook turns moves like these into a 90-day plan with scripts and worksheets, in a Operations & Administration edition. See what's inside
Asked about this role family.
How does AIProof score a role that isn't listed?
The unlisted option carries the model's midpoint weighting, 8 of 15. Your remaining seven answers (routine share, AI overlap, company posture, judgment share, AI usage, restructuring signals, industry) determine the score from there.
Are hybrid roles safer than specialized ones?
Sometimes. Roles that mix judgment, relationships, and varied work resist automation better than single-task roles. But a hybrid of three routine tasks is just three exposures sharing a desk. The task mix decides, not the hybridity.
Which role page should I read instead?
Read the family closest to where your hours actually go. If half your week is reporting, read the data entry and operations pages. If it is client work, read sales. Your biggest time block is your real role family.
Eight questions. One is about your role.
The other seven decide your number.
Hybrid & Unlisted Roles sets 8 of 109 possible points. Your routine share, AI usage, and company posture set the rest. Free, 3 minutes, no signup.
Score Your Actual Risk