AI, jobs, and management: panic less, prepare more
Plenty of smart people are worried about AI hitting the jobs market hard. Jerome Powell is watching it closely, Anthropic's Dario Amodei puts potential losses for entry-level white-collar roles near 50%, and Geoffrey Hinton warns of massive unemployment. No surprise that a recent Pew Research report found 52% of workers are concerned about AI's long-term impact at work, and a third expect fewer opportunities.
On the other hand, Bank of America's Brian Moynihan isn't buying the doom loop. He points to history: in 1969 the U.S. had roughly 80 million people working; by 2019 it was about 160 million. Through waves of "this time it's different" technology, employment doubled. Managers didn't vanish. Manufacturing didn't, either.
What top bankers are actually saying
Moynihan's view is balanced: AI will be disruptive, across every layer of banking, but economies have eaten big tech shifts before. His point is practical-AI augments people, compresses "flight time," and boosts knowledge acquisition. That matters for auditors, lawyers, investment bankers, and, yes, senior leaders too.
Jamie Dimon strikes a similar chord. He sees a future where people may "work less hard but have wonderful lives," while being blunt that "It will eliminate jobs … people should stop sticking their heads in the sand." His push: have real plans to save jobs where possible, retrain people at scale, or use early retirement thoughtfully-because dropping someone from $150k to $30k overnight is a fast path to social blowback.
Reality check from the field
Big banks aren't waiting. BofA is already using automated models and algorithmic tooling in trading, and ranks in the top 10 of the latest banking AI index by Evident AI. JPMorgan leads the group, with Capital One and Royal Bank of Canada close behind. Translation: the adoption curve is steep, and the bar for productivity is moving now.
The signal for managers
Both things can be true: AI will remove some jobs and create new ones. It will also strip tasks from many existing roles and compress skill ramps for junior talent. That's upside if you plan for it-and a mess if you don't. The work you do in the next six months determines whether your team benefits or gets blindsided.
A practical playbook you can start this quarter
- Run a task audit: list recurring workflows by team, frequency, time spent, and risk level. Target high-volume, medium-risk tasks first.
- Pilot "copilots" in role-specific flows: drafting emails, summarizing calls, writing first-pass analyses, QA checklists, and knowledge search.
- Redesign roles for AI-era output: define which tasks are automated, augmented, or remain human-led. Update job descriptions and goals.
- Build skills ladders: pair juniors with AI for speed, then rotate into higher-complexity work so career paths don't stall.
- Create a reskilling track for at-risk roles: 8-12 week programs that move people into analytics, operations, compliance, or customer success.
- Decide now on early retirement and redeployment policies, before a reorg forces your hand.
- Add controls: human-in-the-loop approvals, data loss prevention, and model usage rules tied to risk tiers.
- Measure time saved, not just headcount reduced: cycle time, case throughput, defect rate, CSAT, and revenue-per-employee.
- Stand up an AI governance group with legal, risk, security, and HR. Clarify who can ship what, and on what data.
- Communicate weekly: what's live, what's next, what changed, and why it helps your people.
Guardrails that actually matter
- Data: restrict sensitive fields, log prompts and outputs, and keep customer data in approved environments.
- Quality: document known failure modes, require source citations for high-stakes content, and sample outputs for bias and hallucinations.
- Access: tiered permissions by role and use case; separate testing from production.
- Vendors: review model and tool contracts for IP, indemnity, uptime, and auditability.
Metrics to manage by
- Time-to-first-draft and time-to-close for core processes.
- Error and rework rates pre/post AI.
- Customer response times and satisfaction.
- Training hours per FTE and percentage of work automated/augmented.
- Internal mobility: percent of at-risk employees successfully redeployed.
Workforce planning you can defend
Entry-level white-collar work will feel the squeeze first. Keep the talent pipeline healthy by pairing junior roles with AI-enabled apprenticeship and faster rotations. For senior roles, expect scope expansion-fewer managers, wider spans, clearer operating metrics. If reductions are unavoidable, phase changes with real reskilling and fair exits.
Upskilling that sticks
Don't send people to random tutorials. Map learning to roles and workflows, ship templates, and coach to outcomes. If you need a structured path by job function, see this catalog of role-based programs: Complete AI Training - Courses by Job.
Bottom line
AI will cut some jobs. It will also make good teams frighteningly effective. Managers who redesign work, protect their people with guardrails, and measure real outcomes will come out ahead. The window to act is open right now-use it.
Your membership also unlocks:
