AI Now Decides Promotions and Layoffs for 65% of Managers—Here’s How to Thrive Anyway

Let’s get something straight: the robots aren’t coming for your job—they’re already sitting in the meeting room. A fresh ResumeBuilder survey of 1,342 U.S. managers (released July 2, 2025) says 65 percent of managers are leaning on AI to decide who gets promoted, who gets a raise, and who gets the boot. In some teams, the algorithm is literally the tie-breaking vote. Whoa.

As a workplace strategist who spends his days helping professionals protect their time and energy, I’ve read plenty of “AI apocalypse” headlines. Most are noise. But this one matters because promotion, pay, and job security touch every ounce of our motivation and mental health at work.

“It’s essential not to lose the people in people management.” — Stacie Haller, chief career adviser, ResumeBuilder

I couldn’t agree more. Below is my action-oriented playbook for employees and leaders who want to stay in control while the software crunches the numbers.

1. Understand What the Algorithm Actually Sees

Most AI HR tools ingest performance metrics, project trackers, feedback comments, email sentiment, and even calendar metadata. Many do not weigh soft skills that happen off-platform, like mentoring a new hire over coffee or calming down a panicked client on the phone.

• Map Your Data Footprint. Make a quick list of every system that captures your work: task board, CRM, customer chats, code repo, LMS, etc. That’s your “input layer”—treat it like your portfolio.
• Close the Soft-Skill Gap. If you manage people, bake qualitative feedback into workflows. I like a monthly “wins & wise moves” form where peers note actions the metrics missed.

2. Write for Two Audiences: Humans and Machines

When you post project updates or performance notes, imagine a future algorithm scanning that text. Use crisp subject lines (“Q3 launch blocked—vendor delay 3 days”) and verbs that scream impact (“cut costs by 18%,” “resolved 42 support tickets”). Yes, it feels robotic—but remember, the human reading your update on Friday might be trained by the machine’s scoring on Monday.

3. Make Yourself Algorithm-Proof with “Relational Capital”

Here’s the good news: no model has mastered trust. The fastest promotions I see still come from cross-function reputation.

• Deliver “unscheduled value.” Surprise-and-delight gestures—fixing a team-share spreadsheet before anyone asks, or sharing a custom tutorial—rarely travel through formal dashboards, but humans rave about them in calibrations.
• Network laterally. A 15-minute sync with an adjacent team each week plants eyewitnesses who can override questionable AI verdicts.

If you’re a manager, institutionalize this: host quarterly “story-slam” meetings where team members spotlight a peer’s quiet win. It pumps culture and gives you rich qualitative data to feed back into the system.

4. Demand (and Provide) Transparency

The survey also revealed that two-thirds of managers using AI to manage people have no formal training. That’s a lawsuit waiting to happen. Whether you’re leading a division or an individual contributor, ask the simple question:

“Which decision points in our talent process are influenced by AI, and how do we audit for bias?”

If you’re the one implementing the tool, publish a one-pager that outlines:

• Data sources the model ingests
• Weightings or scoring logic in plain English
• Human override procedures
• An appeals pathway (yes, like grade school—it works)

The point isn’t to slow innovation; it’s to keep morale (and legal exposure) intact.

5. Strength-Train the One Skill AI Can’t Replicate: Judgment

AI can surface patterns; it can’t weigh context with empathy. Make judgment your signature move:

• Run “what-if” drills. Once a quarter, take a complex decision you made and reverse-engineer the factors. This mindfulness reps the muscle algorithms lack.
• Use AI as an advisor, not a boss. Ask the tool to provide three scenarios, then you choose—explain your reasoning in a quick Loom or Slack thread so your thought process is documented.

6. Build a Personal “AI Ethics Clause”

The most confident professionals I coach have a non-negotiable line: they won’t participate in opaque, potentially biased decisions. Draft your own clause in a paragraph, keep it in your desk, and reference it when policy debates arise. A clear moral compass speeds up tough calls and earns respect up the chain.

7. Optimize Energy, Not Just Output

Finally, remember why we chase productivity in the first place—to free capacity for the work (and life) that lights us up. Off-load automatable tasks to the same AI that’s judging you, then spend the saved hours sharpening relationships, creativity, and strategic thinking. Those are the currencies that survive every tech wave.

Bottom line:

AI is now a full-fledged stakeholder in your career. Treat it like any stakeholder—understand its goals, speak its language, and counterbalance its blind spots with uniquely human strengths. Do that, and you’ll not only survive the 65 percent trend—you’ll use it as a springboard.