How the Mandlys AI job risk test works: methodology and limitations
By the Mandlys editorial team · Updated
The Mandlys AI test estimates how exposed your work is to AI and automation, expressed as a future index from 0 to 100. The index is based on your occupation, how your working time splits across types of tasks, your role and how you develop your skills. It is a self-assessment and an estimate, not a prediction that you will lose your job.
What the test measures
The test measures exposure: how much of your work consists of tasks that AI and other automation can already perform or support. Exposure is not the same as a job disappearing. Research distinguishes between tasks that can be fully automated, tasks where AI augments people and occupations that are barely affected. See AI exposure vs job replacement.
What you tell us
You answer questions about:
- Occupation: choose from about 30 occupational groups or enter your own.
- How your time is split across four task types: standardised and rule-based, analytical, relational (people, customers, leadership) and creative or strategic.
- Seniority and decision authority.
- Customer proximity: how much of the work happens in direct contact with customers or clients.
- Skills development: how you learn and how much time you spend per week.
No account is needed. We do not ask for your name. Free-text fields and email are optional.
How the score is calculated
- Occupation baseline. Each occupational group has a baseline exposure between 0 and 1. These values are Mandlys' own estimates, informed by published research on automatability by occupation, such as Frey and Osborne (2017) and the task-based analyses from the OECD and the ILO.
- Task adjustment. A high share of standardised tasks raises exposure. Relational and creative tasks lower it, in line with task-based research showing routine, rule-based tasks are easiest to automate (Autor, Levy and Murnane 2003).
- Role adjustment. Higher seniority, more decision authority and close customer contact lower exposure slightly.
- Development adjustment. Regular skills development, especially in AI and new technology, lowers exposure slightly.
- Future index. The adjusted exposure is capped between 0 and 1 and inverted: 100 means the lowest estimated exposure. 70 or above counts as low risk, 40 to 69 as medium and below 40 as high.
The calculation is deterministic: the same answers give the same result.
What the full report adds
The free result shows the future index and a short summary. The full report ($4.90) adds a task-level walkthrough, risks and protective factors, adjacent roles and a 90-day plan. Parts of the report text are written by an AI language model based on your answers. It is guidance, not a guarantee or expert opinion.
What the test cannot predict
- Whether you personally will lose your job. That depends on employers, the economics of your industry, regulation and how fast technology is adopted.
- Exact timing. The time horizon shown is a rough band, not a date.
- Local conditions. The model does not account for individual employers or regional labour markets.
- New jobs. IMF and ILO research suggests AI both replaces and augments work. The test mainly shows exposure and where you can strengthen your position.
The index is not a scientifically validated probability of job loss. It is a structured way to reflect on your work.
💡 Want to see how this applies to your own job? The free test takes about five minutes and needs no account. → Take the free AI test
Data and privacy
Results are stored anonymously and accessed through a unique link that only you receive. Payments are handled by Stripe and Mandlys never sees card details. See the privacy policy.
Review
This methodology was last reviewed on 2026-10-27. Baselines and texts are updated when new research from bodies such as the ILO, OECD or IMF is published. Questions can be sent via the contact page.
Mandlys is not affiliated with or endorsed by any of the organisations cited.
Frequently asked questions
Is the score a probability that I will lose my job?
No. The future index estimates how exposed your work is to automation, based on your answers. It is not a probability of job loss.
What research is the test based on?
Baselines and weights draw on task-based automation research, including Frey and Osborne (2017), Arntz et al. for the OECD (2016), the ILO generative AI exposure indices (2023 and 2025) and the IMF (2024). Values per occupation are Mandlys' own estimates.
Do I need to share personal data?
No. No account is required. Email and free text are optional.
Sources
- Frey & Osborne (2017). The future of employment: How susceptible are jobs to computerisation? Technological Forecasting and Social Change, 114.
- Arntz, Gregory & Zierahn (2016). The Risk of Automation for Jobs in OECD Countries. OECD Social, Employment and Migration Working Papers No. 189.
- Autor, Levy & Murnane (2003). The Skill Content of Recent Technological Change. Quarterly Journal of Economics, 118(4).
- Gmyrek, Berg & Bescond (2023). Generative AI and Jobs: A global analysis of potential effects on job quantity and quality. ILO Working Paper 96.
- Gmyrek, Berg, Kamiński m.fl. (2025). Generative AI and Jobs: A Refined Global Index of Occupational Exposure. ILO Working Paper 140.
- Cazzaniga m.fl. (2024). Gen-AI: Artificial Intelligence and the Future of Work. IMF Staff Discussion Note SDN/2024/001.
- OECD (2023). OECD Employment Outlook 2023: Artificial Intelligence and the Labour Market.