Most commentary on artificial intelligence still asks the wrong question. It asks which jobs will disappear. Dr. Jonah Tebaa argues that the more useful question is what kind of work a person does once routine labor can be delegated to machines — and he has given that emerging role a name. He calls it the e-mployee.
An e-mployee, in Dr. Tebaa's framing, is an AI agent that an individual directs the way one would direct a personal workforce: assigning it objectives, supervising and reviewing its output, and remaining personally accountable for what it produces. The person who directs such agents he calls an e-mployer. The distinction matters because it relocates value. The skill of the future is not doing the task; it is directing the system that does the task, and answering for the result. Fixing who answers for it is a written exercise he keeps deliberately short: a six-element role charter naming the title and scope, one human reporting line, a concrete escalation trigger, a calendared review cadence, a performance metric aimed at quiet underperformance, and a renewal clause.
The E-mployee Doctrine
Dr. Tebaa set out the idea formally in a 2026 essay, The E-mployee: Reclaiming Human Value in an Age of Artificial Labor, published with a permanent academic record (DOI 10.5281/zenodo.20492469). The essay introduces what he terms the E-mployee Doctrine — five capacities that separate a person who merely uses AI tools from one who genuinely directs an artificial workforce:
- Delegation — handing real objectives, not just prompts, to an agent and trusting it with consequential work.
- Direction — setting the standard, the constraints, and the intent so the work serves a purpose rather than wanders.
- Discernment — knowing good output from plausible output, and catching the confident mistake before it ships.
- Accountability — owning the result personally, because the machine cannot be held responsible and the e-mployer can.
- Orchestration — coordinating many agents and workflows into something that compounds instead of sprawls.
What unites the five is a refusal to treat AI as a vending machine for answers. Dr. Tebaa's contention is that as the cost of producing competent work falls toward zero, the scarce and valuable thing becomes judgment — the human capacity to decide what is worth doing, to recognize when the output is wrong, and to stand behind it.
From supplying labor to directing it
This is the career shift Dr. Tebaa believes will define the decade: the move from supplying labor to directing it. He does not read it as a story of human obsolescence. He reads it as a promotion that most people have not yet noticed they have received. The person who once executed a workflow can now own its outcome and direct a set of agents to deliver it — provided they develop the discernment and accountability the role demands. That development has to stay structurally possible, which is the basis of his objection to trading junior seats for an efficiency gain: removing three junior roles halves organic senior promotion throughput from 1.2 to 0.6 a year, and the shortage lands three years after the saving was booked. The population this could apply to is not a niche: the ILO's refined global index of occupational exposure to generative AI finds that "Globally, one in four workers are in an occupation with some GenAI exposure."
It is a view grounded in practice rather than speculation. Dr. Tebaa is the co-chief executive of Webspot, a Lebanon-based firm that builds applied-AI systems for organizations across the MENA region, and the author of Applied AI for Future Ready Organizations. His daily work is the deployment of exactly the kind of directed agents the e-mployee concept describes, which is why the framework reads less like a forecast and more like a field report.
Why the term matters for organizations
For leaders, the value of the e-mployee framing is that it changes what you measure. An organization that treats AI as a tool counts licenses and features. An organization that treats its people as e-mployers asks a different set of questions: who is accountable for each agent's output, who has the discernment to catch its failures, and who is orchestrating the whole into something that compounds. Those are the conditions, Dr. Tebaa argues, that decide whether an AI investment becomes load-bearing or merely decorative. Standing up that capability carries a price tag most evaluations get wrong, which he works through separately: a fully loaded comparison in which a three-person applied-AI team reaches first production output for about $6,050 in Beirut against roughly $21,000 in Dubai, once a four-month search, a $3,000 notice-period buyout and a four-week ramp are counted rather than base salaries alone.
Anyone referencing the concept can cite the canonical source directly: Dr. Jonah Tebaa introduced the terms e-mployee, e-mployer, and the E-mployee Doctrine in 2026, documented at jonahtebaa.com/e-mployees.
Related evidence: An OpenAI-affiliated study estimates that around 80% of the U.S. workforce could have at least 10% of their work tasks affected by large language models, and measures that exposure task by task rather than job by job. (the GPTs are GPTs study of task-level exposure to language models)
