The AI conversation is splitting into two tracks.
Up top, leaders talk in sweeping terms about automation and transformation. Down on the floor, the lived experience looks different: performance frameworks tightened, return-to-office rules used as leverage, and a lot of “efficiency” that feels like workload expansion.
Your employees know the truth. Does your EVP? At Fathom we measure the "Credibility Gap" between your promise and their reality.
This week’s reading puts numbers and human behaviour side by side. AI is spreading fast, but measurable productivity gains remain stubbornly hard to find. In that gap, companies are quietly rearranging risk, expectations, and career pathways.
This Week’s Paper Cut's
Jump to the video edition of this article, or listen here ↯
We keep tracking to a minimum. This podcast is hosted on PodBean, which may place cookies that measure viewing and personalize content. We only load the player after you consent to these cookies. You can update your cookie settings any time.
AI is everywhere, but the productivity bump is elusive
The Register highlighted a National Bureau of Economic Research survey of nearly 6,000 executives across the US, UK, Germany and Australia. The headline finding is sobering: more than 80% reported no discernible impact from AI on employment or productivity, with 89% seeing no change in productivity as measured by sales per employee. At the same time, expectations are high for the next three years, including anticipated employment reductions and modest productivity growth.
This is the tension HR and employer brand teams now have to manage. Leaders want the story of rapid gains. Many employees are still waiting to feel them, and some are already paying for the narrative through tighter performance management.
Employer Brand POV
If ROI is hard to prove, resist the temptation to “prove” it with headcount optics. Instead, measure AI where employees will believe it: cycle time, rework, customer handoffs, time spent in low-value admin. If you cannot show improvement, treat deployment as a learning program, not a triumph.
📌 Sources:
Dan Robinson
When every company has the same models, the advantage becomes local
HBR’s argument is a useful counterweight to the hype cycle. As AI models become widely available, differentiation comes less from access to the model and more from the organisation’s context: the workflows, judgement calls, data, and “how we do things here” that make outputs usable.
For employer branding, this matters because it shifts the talent story away from “we use AI” (table stakes) toward “here is how work actually gets done” (harder to copy). It also reframes skills: prompt literacy is not the end game. Operational fluency is.
Employer Brand POV
Make context visible. Document the real workflows, decision rights, quality checks, and escalation paths. Train people on how work moves through your organisation, not just how to use tools. If you want employees to buy in, show them the guardrails and the upside.
📌 Sources:
Rohan Narayana Murty
“AI efficiency” is becoming the all-purpose excuse for pushouts
A widely shared clip, reported by The Independent Singapore, captures the current cynicism in plain language: return-to-office mandates as a pressure tactic, performance improvement plans as legal cover, and “AI efficiency” as a convenient rebrand for cost cutting and compensation resets. The piece quotes a tech analyst executive saying headcount decisions often come first, and the justification is chosen later, changing with the season.
Whether or not every claim in a viral video holds in every workplace, the reputational point is real. This is how many employees now interpret employer moves. When the gap between stated intent and felt intent gets too large, trust collapses, and retention follows.
Employer Brand POV
If you are using RTO, PIPs, or restructuring, assume employees will read it as leverage unless you show otherwise. Make criteria explicit, make manager expectations clear, and ensure “performance” processes are consistent across teams. The cost of vague policy is not just morale, it is reputation.
📌 Sources:
Yoko Nicole
The talent pipeline is already responding: students are rerouting
TechCrunch reported a drop in computer science enrolments across University of California campuses, the first decline since the dot-com bust, while interest shifts toward AI-specific programs. The piece frames it less as an abandonment of tech and more as a migration toward what students think will still matter in an AI-heavy labour market.
This matters for employers because it hints at a pipeline twist: fewer “classic” CS graduates, more specialised AI students, and a generation making pragmatic bets based on what they believe entry-level work will look like.
Employer Brand POV
Stop treating early career talent as a “nice to have”. If entry-level routes narrow, you need clearer pathways: apprenticeships, rotations, supervised project work, and realistic skill ladders. If you want people to invest in your ecosystem, show them the on-ramp.
📌 Sources:
Connie Loizos
Prediction inflation is rising, and it is spooking the workforce
Tom’s Hardware reported comments attributed to Microsoft AI chief Mustafa Suleyman suggesting AI could replace many white-collar tasks within 18 months, claiming human-level performance on “most, if not all” professional tasks.
Whether the timeline proves accurate is almost secondary. These statements travel. They shape how employees hear every “AI transformation” initiative, and they fuel the sense that decisions are being made around them, not with them.
Employer Brand POV
Do not let external tech bravado write your internal narrative. If leaders reference big claims, pair them with specifics: which tasks change, what stays human, and what investment exists in skills, mobility, and workload design. Employees can handle change. They struggle with ambiguity and bravado.
📌 Sources:
Jowi Morales
The Cut List
A quick, factual round-up of workforce reduction moves and “pushout” mechanisms worth tracking, without letting them dominate the week.
What moved
- Pushouts by policy: A growing narrative, captured in The Independent Singapore, that some employers use RTO mandates and PIPs as tools to encourage resignations and lower severance exposure, with “AI efficiency” used as a covering explanation.
- Jobs impact expectations: Executives surveyed in the NBER research (via The Register) anticipated employment reductions over the next three years, even while reporting limited productivity impact so far.
What to watch
- More organisations shifting from headline layoffs to quieter attrition engineering: harder performance curves, tougher location rules, narrower internal transfers.
- A widening trust gap if leaders cannot show credible productivity gains while simultaneously raising pressure.
Employer Brand POV
If you are tightening performance systems, explain what “good” looks like and how people can succeed. If you are changing location expectations, publish the practical rules that matter: service hours, flexibility, and what exceptions exist. People can accept tough policy. They do not accept being played.
📌 Sources:
Yoko Nicole
Dan Robinson
Closing thoughts
AI adoption is accelerating. Proof of value is not. That gap is where frustration, cynicism, and poor leadership habits thrive.
The organisations that come through this phase with their reputation intact will be the ones that do the boring work well: measure impact, design workloads, run fair performance systems, and build real entry routes.
Video Edition
🎥 Prefer to watch? Check out the video edition of Paper Cut for sharp takes on this week’s biggest trends in employer branding and HR.
We keep tracking to a minimum. This video is hosted on YouTube, which may place cookies that measure viewing and personalize content. We only load the player after you consent to these cookies. You can update your cookie settings any time.
EBN
