The FT asks what happens to managers' grades: check the headcount pay yardstick, not a proven pay cut

On Nov. 8, 2026, the Financial Times published a column by Pilita Clark under the headline "AI is coming for managers' headcount." The dek is "Their pay grade could be next." The paper's official account forwarded the same link that day and tagged it as opinion.

The Straits Times ran a readable opening the same day, with the title "How do you pay bosses of AI agents?" Clark writes that she was chairing a panel on artificial intelligence when someone in the audience asked about something she had never thought of before. The way an average manager is graded and paid often depends on how many people they manage. The higher the headcount, the higher the pay and seniority. What happens when a manager's human underlings are replaced by AI agents? What a reader can check is that question, and the dek's suggestion that pay grade could be next in line. Not a finding that companies have already cut managers' pay bands.

Here is what I actually read.

The FT's management index and Clark's author page list the column on Nov. 8, 2026, with the headline and dek above, and name Pilita Clark as the writer. The paper's account on X forwarded the link at 06:59 UTC that day. The post text is the headline, plus the word opinion. The original FT page returned a security-verification screen in this environment and did not yield the body.

The Straits Times published its version at 17:12 Singapore time on the same day, which is 09:12 UTC, under the title "How do you pay bosses of AI agents?" The byline is Pilita Clark. The readable dek says workplace AI is challenging traditional yardsticks of compensation for managers based on human headcount. Another line says research shows "AI brain fry" is an occupational hazard for managers of teams of bots. Then comes Clark's own opening: "I was chairing a panel discussion about artificial intelligence the other day when someone in the audience asked about something I had never thought of before." The question follows: "The way your average manager is graded and paid often depends on how many people they manage. The higher the headcount, the higher the pay and seniority. But what happens when a manager's human underlings are replaced by AI agents?" The page then asks readers to continue under a subscription wall. I do not have more of her sentences.

A summarizer that cites the FT column also listed a McKinsey percentage, comments attributed to Kit Krugman of Altana, a warning about "AI brain fry" from Brian Elliott, and a large rise in "responsible AI" job ads in banking. I did not see those names or figures in the Straits Times opening or on the FT page itself, so I do not use them here.

The headline treats headcount as the thing under pressure. The dek treats pay grade as something that could be next. That is the column's framing, not a completed pay survey. The test is whether employers later rewrite the rules they use to grade managers: still by the number of people they oversee, or by the work, the risk, or the software agents they supervise. If pay bands do not move, and only job titles do, the word "could" in the dek remains a conjecture.

The readable Straits Times opening puts the question in concrete terms. Grading and pay "often" depend on how many people a manager oversees. More people have meant more pay and more seniority. The live question is what happens when those people are replaced by AI agents. Clark presents this as a question put to her by an audience member, not as a formula she has already copied out of company handbooks. That limits what the visible text can support. What we have is a challenge to a common yardstick, not a table of new salaries.

"AI brain fry" appears in the Straits Times presentation of her argument, which says research shows it is an occupational hazard for people who manage teams of bots. The visible body does not name the study, the sample, or the authors. Until the full column is readable, that phrase should not be treated as a clinical finding the piece has already pinned down.

The strongest steelman is this: paying managers by headcount assumes that each extra person adds coordination, review, and responsibility. If software agents take on scheduling, reporting, and the passing of information, that yardstick fails in both directions. A manager with few human reports and many agents can look smaller on an organization chart while the workload rises. If agents are counted as part of team size, a grade can also inflate without any increase in human staff. The dek's line that pay grade "could be next" can point either way. It would be supported if large employers publish grade manuals that distinguish human reports from software agents. It would be weakened if firms keep using the old headcount formula and treat agents as ordinary tools.

The risk is the missing middle of the column. I did not read the full FT text, so I cannot import names and percentages from a summarizer and treat them as evidence Clark has already checked. A column that opens with a question is also not proof that managers' pay has already been cut across the board. Reading "could be next" as "has already happened" is a step the headline does not support.

What can be watched in the coming months is whether large employers publish rewritten grade criteria for managers that say how software agents count; whether pay consultants or union documents start asking if supervising agents should be priced on its own; and whether "AI brain fry" moves from a column phrase into occupational-health or human-resources guidance with a definition attached.

A reader should score this story against grade manuals and pay policies, not against the already sharp headline that AI is coming for managers' headcount.

What can be tested in Clark's column is that managers are often graded and paid by the number of people they manage, and that this yardstick is now in question if AI agents replace those reports. Not a finding that pay bands have already been cut. If later corporate rules match that question, this is an argument about how management work is priced. If they do not, the piece remains a comment that used headcount as a sharp title.

【No.2】