DWCLA cognitive load assessment

How much the work is asking of a head: eight statements over three sections, and the index that reads them against how often the hour is interrupted.

  • 1

    Eight statements

    Each one is rated from 1 to 5. One is worded the other way round, and the page handles that.

  • 2

    Your switches

    How often an average hour is interrupted — the second half of the index.

  • 3

    A number to take with you

    A load score from 1 to 5, an index from 1 to 25, and the section that is carrying most of it.

The tool

Where the argument turns into a number.

Each statement is rated from 1 to 5. Honesty is more use here than hope, and the answers stay in this browser.

1 = strongly disagree · 5 = strongly agree

1Work organisation ISO 45003
A1I find it difficult to stop thinking about technical problems after working hours.
A2I feel that AI tools have raised the expectation for speed beyond my mental capacity.
A3I rarely feel a sense of full closure on my tasks at the end of the day.
2Human–AI interaction ISO/IEC 42001
B1My role feels ambiguous: I am not sure whether I am a maker or a reviewer of what AI makes.
B2I worry that my core skills are wasting away because I lean on AI too much.
B3The fear that AI will make something up keeps me on edge.
3Social factors ISO 45003
C1I spend more time with AI and screens than with the people I work with.
C2I feel safe saying “I don’t know” without fearing the comparison with AI.
Worded the other way round, so agreeing is the healthy answer.
SContext switches
How many times an hour does something pull you off task?
Messages, meetings, alerts, someone at your desk. What counts is an average hour, not the worst one.

What this is, and what it is not.

This is a pilot screening tool, not a clinical instrument, and it has not been statistically validated yet: validation on samples across sectors is in preparation. It tells you where to look and what to talk about; it diagnoses nobody. If the answers worry you, that is a conversation to have with a person, not with a page.

Take part in the validation

The pressure

We hardened everything except the processor that does the thinking.

Firewalls, zero trust, intrusion detection: we spend on all of it, and leave the human mind open to a flood of information and worry that never stops arriving. Three pressures do most of the damage.

  • The efficiency paradox

    The breaks were the work we gave away

    The routine tasks are gone, and with them the quiet minutes a mind used to rest in. What is left is decision after decision, each one high-stakes. The day is shorter on work and longer on strain.

  • The open loop

    The work is never finished, so it never closes

    An unfinished task stays active in the head. With continuous deployment, a phone that never sleeps and a backlog without a bottom, the loop stays open long after the laptop is shut, and it feeds on the hours meant for rest.

  • Cognitive debt

    A tired mind is a security and quality risk

    Fatigue does not stay in the person. It shows up as the click on the phishing link, as the defect that reaches production, and as the resignation that takes years of institutional memory with it.

What it costs

Burnout is an operating expense before it is a feeling.

It arrives as breaches, as defects, and as people who leave. Three numbers from outside this page, for an internal audit to be read against.

68%

of breaches involve a non-malicious human element

Verizon DBIR, 2024

77%

of employees say AI tools added to their workload

Upwork Research Institute, 2024

0.5–2×

annual salary to replace one employee who leaves

Gallup, 2019

The controls

Three rules, laid over the human operating system.

Each one answers a named risk, and each one is written so a manager can put it in place on Monday rather than agree with it on Friday.

  • A

    The hard stop

    A right to disconnect that the tooling enforces, and fifteen minutes before logout to write down what is still open. What is written down, the brain can sleep on.

    Answers: Cognitive overload · the open loop

    • ISO 45003
  • B

    Human in command

    The hierarchy written down: the human is the architect and the reviewer, the model is the drafter. Juniors try the problem by hand before they ask, so the skill survives the help.

    Answers: Displacement anxiety · impostor feeling

    • ISO/IEC 42001
  • C

    Social recovery

    Breaks away from the screen, and sessions where the senior people tell the room what the model got wrong and what they got wrong. Safety is built by whoever admits first.

    Answers: Digital isolation

    • ISO 45003

Where the authority comes from

Not wellness tips. Three standards, read together.

Wellbeing sits with human resources, the models sit with technology, and the gap between the two departments is where the harm collects. The protocol closes it by reading three international frameworks as one.

  • ISO 45003:2021

    Psychological health and safety

    Guidance inside ISO 45001 for managing psychosocial risk at work.

    Application: It is what lets you say that human capacity sets the pace of the work, not the speed of the model.

  • ISO/IEC 42001:2023

    AI management systems

    The management system for artificial intelligence inside an organisation.

    Application: Psychological harm is written into the risk assessment, and human oversight is written into the job.

  • IEEE 7010-2020

    Wellbeing metrics

    A recommended practice for measuring what intelligent systems do to human wellbeing.

    Application: It is the move from feeling to measuring: agency, competence, and the index this page calculates.

About the assessment

From my white paper The Cognitive Firewall (v1.1, 2026), Appendix A, which sets the eight statements and the thresholds used here.

How to cite

Elgendi, M. F. (2026). The Cognitive Firewall (White Paper, Version 1.1), Appendix A: Digital Workforce Cognitive Load Assessment. www.fawzooz.ai