Anthropic's CEO is dining with a president on the other side of the AI debate. The morale question it implies.
On September 27, 2026, TechCrunch reported that Anthropic CEO Dario Amodei "seems to be everywhere this weekend": lampooned on the season premiere of Saturday Night Live, and set to have dinner with President Donald Trump at the White House that evening. Axios first broke the news of the dinner, which TechCrunch said it confirmed with "a source familiar with his plans." According to TechCrunch, it will be the first one-on-one meeting between two men who "recently found themselves on opposite sides of the AI safety debate." The friction predates that. TechCrunch noted that earlier this year the Pentagon designated Anthropic a supply-chain risk after the company tried to put guardrails around the use of its technology, a designation Anthropic has been fighting in court, "although other administration officials have been friendlier." The article does not say what the two will discuss.
Set the politics aside and look at the shape of the weekend: a leader mocked in public, a legal fight still unresolved, and a decision to sit down across the table from the other side of the argument anyway. The report says nothing about how Anthropic's staff feel, and this essay makes no claim about it. But most managers have lived through a smaller version of that weekend. The team has been criticized, overruled or made the punchline, and the problem won't resolve on any timeline the manager controls. In that stretch the team isn't watching the outcome. It's watching the leader's posture, to see whether they hide, spin, or keep engaging. Morale rests on that, and it's why a question that sounds soft is one of the harder ones in a management loop.
Why they ask it
Morale is a lagging indicator of management, and interviewers can't observe it directly. Attrition and quiet disengagement usually trace back to a stretch when the team sensed something had gone wrong and nobody with authority named it. The question tests three things. Did the candidate notice the shift before it reached the metrics? Did they diagnose the cause rather than treat the symptom? And did they change something structural, or only the temperature for an afternoon? For senior roles it also tests composure. The interviewer wants to know whether the candidate holds steady when the cause is outside their control, because that's when a team takes its cues from its manager.
The trap
The weak answer is about an event: a team lunch, an offsite, a shout-out channel. Each is pleasant, and none is a morale intervention on its own, because none touches the reason morale dropped. A candidate who describes a lunch that made people feel more connected has shown that they treat morale as mood.
The subtler trap is the rescue story, where the candidate fixed the underlying problem and morale followed. It sounds strong, but it dodges the question. Morale matters most when the cause can't be fixed quickly, and a candidate whose only move is removing the cause hasn't shown what they do when they can't.
The third failure is a vague result. Saying the team was much happier is an assertion, not evidence.
Applying STAR-T
Situation. Choose a stretch with a named cause the team couldn't wish away. Our platform team spent months on a migration tool, and leadership cancelled it days before launch in favour of a vendor product. Two engineers asked me privately whether their work had been wasted. What people actually said out loud is what makes the situation credible.
Task. State the real job, which is rarely making people happy. I couldn't reverse the decision. My job was to stop the team concluding that their judgement didn't count here, and to keep strong engineers from quietly interviewing elsewhere.
Action. This is where the answer is won, and it needs distinct moves in a clear order. I got the actual reasoning from the VP instead of forwarding the announcement, and I shared it that week, including the parts I disagreed with. I asked the VP to attend our retro and take questions directly. I went through the tool with each engineer and found what carried over, and the validation layer ended up in the vendor integration. Then I gave the engineers who'd raised concerns ownership of that integration. Nothing here is a celebration. Each move answers a specific loss: of information, of voice, of work, of ownership.
Result. Give evidence the interviewer can picture. Both engineers stayed, the integration shipped on schedule, and our team's engagement score on feeling heard held steady while the wider organization's fell. Without a survey, retention, ownership people volunteered for, or a candid remark from a team member all work.
Trade-off. Name the cost. Asking the VP to face a room of unhappy engineers spent capital, and that relationship was cooler for a while afterward. I'd do it again, but it wasn't free. Morale questions reward this, because admitting a cost marks the story as a real decision rather than a feel-good anecdote.
The follow-up that breaks weak answers
The follow-up that separates candidates is some version of what would you have done if you couldn't fix anything at all? The event story has nothing left to offer. The rescue story is exposed, because its only tool was removing the cause.
A strong candidate's original story already contains the answer. Tell the truth early. Keep decision-makers reachable. Keep engaging with the people on the other side of the decision. Give the team something real to own while the larger issue stays open. The interviewer is listening for one idea: morale isn't the absence of bad news. It's a team's confidence that its leader will stay at the table when the news is bad.
Score your answer against the director’s bar
Q: Tell me about a time you helped boost your team morale?
Bank one story where the cause of low morale was outside your control, and rehearse it until the actions come out in order. Try it free →