Is AI Already Deciding Who Gets Fired? How AI Is Changing Your Job
Who is accountable when an AI recommendation affects pay, promotion, or dismissal—and which rights must the employee retain?
Understand where AI already participates in management decisions, why workplace data cannot capture the full human context, and which transparency, review, and appeal mechanisms are needed before an error becomes a career decision.
What to watch for
Key takeaways
AI can process more signals and prepare a recommendation, but a decision about a person still requires an accountable owner who can explain the consequences.
Lawsuits involving Meta employees show that algorithmic processes affect real working conditions, layoffs, and rights—not an abstract future.
The system collects metrics, compares people, and proposes an interpretation. The danger begins when that interpretation is treated as an objective portrait of the person.
The more access AI receives, the more convincing its conclusion becomes. Yet invisible work, exceptions, and the reasons behind changing metrics may never enter the data.
The experiment with ChatGPT Work and Codex shows that the system does more than accelerate one task: it changes what the person does, delegates, and reviews.
When AI receives broad workplace context, control, information, and responsibility shift within the team. Productivity cannot be measured only by counting automated operations.
An employee must know which data was used, who reviewed the recommendation, and how to challenge an error. Once the career decision is made, correcting the system is too late.
What this episode is about
Artificial intelligence is gaining access to workplace data, evaluating performance, and helping managers make decisions. But when an algorithmic conclusion affects pay, promotion, or dismissal, a question of efficiency becomes a question of responsibility.
AI Is Already Participating in Management Decisions
This is not a distant future in which a robot independently dismisses an employee. AI can already collect performance data, analyze files and applications, compare workers, build rankings, and present a ready-made interpretation to a manager. The more access these systems gain to the work process, the more strongly their conclusions shape how a company sees an individual person. The problem begins when a score or ranking is treated as objective truth.
Data Does Not See Exceptions or Human Context
An algorithm can process more signals than a manager, but it sees only what has been converted into data, rules, and permitted options. It may not know why a metric changed, what invisible work an employee performed, or why a specific situation requires an exception. A person can also be biased and wrong, but a person can doubt a conclusion, hear an explanation, and assume moral responsibility for a decision. A model cannot carry that responsibility.
AI Changes the Way Work Is Organized, Not Just One Function
Anna’s experiment with ChatGPT Work and Codex shows another shift. Once a system gains access to the work context, it does more than accelerate writing or spreadsheet analysis. It changes the sequence of actions, the distribution of roles, and which decisions a person makes independently. The impact of AI therefore cannot be measured only by counting automated tasks; the structure of the work process itself changes, and so does the degree of control over the person inside it.
Rules Are Needed Before an Error Becomes a Career Decision
Lawsuits involving Meta employees and Google workers’ demands for protections during layoffs show that the conflict has already moved beyond a technical experiment. People need to know which data a system uses, whether its conclusion can be challenged, who reviews the recommendation, and who is responsible for an error. AI can be a powerful analytical tool, but handing it decisions about a person’s future without a transparent process is no longer a productivity question. It is a question of what rights an employee retains inside a company that manages work through algorithms.
AI can help a manager see more, but it must not turn incomplete data into an impersonal verdict. Managerial accountability remains human and must be visible to the employee.
Episode transcript
The episode is in Russian; below is an English reading guide to the transcript (the full EN transcript is a machine translation). Voice matching applied to 34 segments: 34 identified, 0 mixed, 0 marked with ✓, and 0 unresolved.
Loading…