The phrase 'AI job replacement risk' suggests a binary — either your job will be replaced or it won't. The reality is more granular. What AI replaces first is not jobs but tasks: the structured, rule-bound, information-processing parts of any role. Whether that task displacement adds up to job displacement depends on what percentage of your working hours those tasks represent, how fast your industry is adopting the tools that automate them, and whether the remaining tasks are enough to justify a full-time position.
A meaningful AI job replacement risk assessment has to work at the task level, not the job title level. The Fossil Score method evaluates five dimensions — task automation exposure, skill transferability, physical presence, relational complexity, and AI literacy — to produce a score from 0 to 100 that reflects how likely your specific role is to survive the current wave of automation. Unlike occupation-level databases that give you a single automation probability for your job title, this method accounts for what you actually do each day.
What the research says
92M
jobs displaced by 2030
WEF Future of Jobs 2025
170M
new roles created by 2030
WEF Future of Jobs 2025
41%
of employers plan AI-driven headcount reductions
WEF 2025
55K
job cuts explicitly attributed to AI in 2024
Challenger, Gray and Christmas
Task automation exposure covers the percentage of your daily tasks that current AI can handle reliably — not hypothetical future AI, but tools deployed today. Skill transferability measures how easily your skills translate to roles that are growing rather than contracting. Physical presence captures how much of your work requires you to be somewhere specific, since embodied work is harder to automate than digital work. Relational complexity reflects how dependent your role is on human trust, ongoing relationships, and interpersonal judgment that clients and colleagues will not delegate to an algorithm. AI literacy measures whether you are positioned to use AI tools to expand your own capacity rather than compete against them.
The interaction between these dimensions matters. A role with 70% task automation exposure but high relational complexity and AI literacy is in a very different position from a role with the same automation exposure but no human-facing work and no AI tool adoption. The assessment weights these dimensions to reflect where the actual risk is concentrated.
Task automation exposure
What percentage of your daily tasks can current AI handle reliably — not future AI, tools deployed today
Skill transferability
How easily your skills translate to roles that are growing rather than contracting
Physical presence requirement
How much of your work requires you to be physically somewhere — embodied work is harder to automate
Relational complexity
How dependent your role is on human trust and relationships that clients won't delegate to an algorithm
AI literacy
Whether you're positioned to use AI to expand your capacity rather than compete against it
A proper assessment measures five dimensions: task automation exposure, skill transferability, physical presence requirement, relational complexity, and AI literacy. Generic job-title lookups only estimate task automability and ignore the factors that make your specific situation meaningfully different from the average person with your job title.
Two people with the same title can have very different risk profiles depending on what they actually do each day. A financial analyst who spends 80% of their time advising decision-makers has a different risk profile from one who spends that time running standard reports. Job-title averages can be significantly wrong for any specific person.
On the Fossil Score scale, scores below 40 indicate high replacement risk (Brachiosaurus classification). Scores between 40 and 65 indicate moderate risk. Above 65 is relatively safer. The score reflects survival probability — a score of 30 means significant risk to the role, not that only 30% of tasks are automatable.
High-risk roles share three traits: tasks are structured and rule-based, outputs are digital rather than physical, and the work involves processing information rather than generating new judgment. Data entry, insurance claims processing, bookkeeping, and standard document processing roles rank highest. Roles with physical presence, complex human relationships, or novel judgment requirements are significantly lower risk.
Yes. A bookkeeping clerk at a firm using Xero with AI features enabled is in a meaningfully different position from one at a practice still using spreadsheets — not because the tasks differ, but because AI is already doing the automatable work at one and not the other. Both face the same eventual pressure, but on different timelines.
Related topics
Free Assessment
Five dimensions, your specific tasks, a Fossil Score from 0–100. Free. Takes 4 minutes.
Get My Fossil Score