AI, Work, and Where to Invest as an Impact VC
- ImpactVC
- Jun 3
- 11 min read
For a recent ImpactVC Deep Dive, we brought together three perspectives at the frontier of AI, labour markets, and what this means for impact focused VCs:
Albert Howard, Head of Research, Revent
Louise Marston, Managing Director, Resolution Foundation / Resolution Ventures
Nick Andreou, Head of AI, Better Society Capital / ImpactVC
The goal wasn’t necessarily to reach consensus, as this is a fast moving and evolving topic. But, it was to sharpen how impact-focused venture investors think about one of the most uncertain and consequential shifts underway: how AI is reshaping work.
What emerged was a set of tensions, signals, and practical implications for impact focused VCs.
Key highlights
The near-term risk might not be mass unemployment, some VCs believe it to be labour immobility - how people move through work when technology reshapes roles, creating a widening friction of people unable to signal their skills, to navigate change, or to access the next set of opportunities as roles evolve.
This wave is structurally different: Previous waves of automation disproportionately targeted routine, manual, or highly structured work. Today's AI systems encroach on cognitive and creative tasks that were once considered safely human. This does not mean those jobs disappear - but it does mean their internal task mix changes.
Career navigation, skills evidencing, and workforce transition tooling are massively underbuilt relative to the pace of change.
AI looks set to remove the bottom 25% of the traditional career ladder - new parallel pathways are urgently needed.
Whether investing in the application layer or infrastructure layer, impact-focused VCs need clearer frameworks to distinguish genuinely beneficial AI from displacement plays - a few tools and resources were listed as starting points.
Signal vs. noise - What is actually happening
"This may be the beginning of a new trend where white-collar jobs become threatened more seriously by AI. Once a few companies start the trend, competitive forces may induce others to follow."
The panel was candid about the limits of current visibility: what is happening is becoming clearer through some examples, but why it is happening remains contested.
There are some key signals observed:
Block cut ~40% of its workforce, over 4,000 employees - CEO Jack Dorsey framed it as a structural AI-driven shift - not a cyclical one, citing that a majority of large companies will reach the same decision within this cycle
Graduate intake reductions are appearing across sectors, not just tech
Over-qualified workers are displacing entry-level candidates, compressing the lower rungs of career ladders
At the same time, UK NEET rates (18–24) are high and rising relative to peer countries - with nearly one million young people aged 16 to 24 in the United Kingdom are not in education, employment or training. The causation is unclear, but employer uncertainty about future skills needs is a plausible driver
The recently released landmark Young People and Work Report (2026) flagged that entry-level roles have become less plentiful and more automated. Apprenticeship starts for young people have declined by over 40%. Young people are being screened by portals, algorithms and recorded interviews before anyone has met them in person.
However, the economic data tells a slightly different story:
According to an Oxford Economics analysis, AI-related job losses in 2025 account for 4.5% of total reported job losses, and job losses attributed to AI are more likely to be overstated than understated.
A 2025 labor analysis from Goldman Sachs determined “aggregate labor market impacts are still negligible,” citing “basically zero” AI contribution to the U.S. GDP growth for the year.
Why this wave might be different
Previous automation waves displaced manual and routine cognitive work. This wave was described to have distinct characteristics that compound the challenge:
AI is reshaping higher-skill work, not just automating routine tasks Previous waves of automation focused on manual and highly structured work. Today’s AI systems are encroaching on cognitive and creative tasks once considered distinctly human.
Jobs aren’t disappearing, but they are being reconfigured Roles in fields like law, finance, research, and software are seeing a shift in task mix. Traditional career pathways where junior talent learned by doing are being disrupted as AI takes on more of this foundational work.
Labour demand is diverging across the economy These shifts are happening alongside persistent shortages in caregiving and trades - sectors where demand is high, automation is harder, and perceptions of “good jobs” lag behind reality.
Perception gaps are creating friction and opportunity There remains a disconnect between the quality and remuneration of certain roles and how they are viewed by workers, parents, and educators. Companies like Greenworkx (reskilling into the green economy) and TaskHer (improving representation in trades) are addressing these perception barriers - and point to a clear investment opportunity.
AI is accelerating demand for skilled trades As Randstad highlights, AI can’t build data centres. Demand for skilled trades is growing up to 3x faster than for professional roles - a reminder that the future of work will be shaped as much by physical infrastructure as by digital innovation.
The pace of change is faster - and potentially harder to absorb The speed and simultaneity of disruption across sectors exceeds previous technology waves, leaving less time for labour markets to adjust organically.
Some jobs (and people) are at greater risk than others.
Consider the following graph from a labor market study by Anthropic, which shows the job categories where AI could have the most disruptive influence (the blue shape) and where AI is currently used (the red shape). Areas like construction, and healthcare support are both growing in the number of jobs and is low-risk to AI disruption.
The reallocation shock
The dominant public narrative frames AI as a job destruction story. The panel's view was more nuanced: the more immediate and less-discussed risk is labour immobility.
"The narrative of ‘AI is taking your job’ places the burden in the wrong place. The real problem is that people get stuck - they can't see where the next opportunities are."
Albert Howard, Revent
AI-driven capability is accelerating rapidly, but labour market transitions lag behind. Workers aren't necessarily replaced - they get stuck. This has compounding consequences: unemployment feeds low productivity, increased fiscal burden, and significant costs to individual purpose and welbeing. Albert dives into this in depth in his piece, The Next Great Reallocation of Work.
Hiring is the bottleneck and signalling is broken
Hiring is the mechanism through which people move from where they are to where opportunity exists. Yet the hiring stack has changed remarkably little relative to the pace of change in work itself.
Today’s system is fragmented and low-signal: recruiters compete for placement, candidates are flooded with irrelevant outreach, and roles are often poorly defined. The result is a process that is noisy, inefficient, and increasingly misaligned with how work is actually evolving.
At the same time, hiring sits at the centre of several intensifying pressures:
Widespread labour shortages across Europe, particularly in construction, engineering, healthcare, ICT, transport, and warehousing
Structural constraints including ageing populations, border friction, and persistent skills mismatches
Economic drag, with estimates suggesting GDP in advanced economies could be up to 1.5% higher if vacancies were filled
These challenges are uneven. In many skilled trades, strict certifications limit flexibility, while in knowledge work, the problem is often the opposite: too much noise, not enough clarity. Platforms like LinkedIn, Indeed, and Totaljobs are increasingly saturated with low-quality listings, making it harder for both employers and candidates to identify real fit.
At its core, this is no longer just a matching problem - panellists highlighted this as a signalling problem.
Candidates struggle to demonstrate capability beyond credentials
Employers struggle to define roles in terms of real skills and outcomes
Both sides lack tools to translate potential, preference, and ambition into meaningful work
A new generation of companies is beginning to rebuild this layer of the market:
Welcome to the Jungle - improving job quality through structured employer and role insights
Breakroom - surfacing frontline job quality through employee-generated data
Forage - enabling career exploration through simulations
Jack and Jill - combining AI-driven recruiting with career coaching
Asap.work - addressing acute shortages in construction through a two-sided marketplace
Greenworkx and TaskHer - tackling access, reskilling, and representation in underserved talent pools
These models point to a shared insight: hiring is not just about filling roles. It is about building better systems to translate skills, preferences, and ambition into work people can access - and actually want to do.
Worker rights, power and autonomy
A recurring point was that worker voice matters in how AI gets deployed. In previous waves of tech adoption, outcomes were better when workers had some agency in the process: adoption was higher, resistance was lower, and the technology was more likely to be used well. The panel noted that there are often parts of work you might think can be automated out, but which actually carry relational or qualitative value when humans are involved.
The Klarna customer service example was cited as a cautionary case: after heavily promoting AI customer service, the company later brought humans back, with reporting citing lower service quality and customer dissatisfaction as reasons. The lesson is not that AI failed, but that a purely cost-driven rollout can destroy part of the value that customers actually pay for.
The broader implication is that impact investors need to understand not just the productivity upside of AI, but also the distribution of power, autonomy, and control in the workplace.
Investment opportunities - some areas the speakers explored for impact-focused VCs to dive into in AI and future of work
Area | Opportunity | Startup Examples | Evidence points |
A. Labour Mobility & Career Infrastructure | The panel explored the strong opportunity in sector-specific hiring platforms, skills evidencing and certification, career transition tools, and infrastructure for independent work, solopreneurs, and the creator economy. The matching, mobility, and navigation layer remains dramatically underbuilt, especially for people moving into sectors such as construction, logistics, healthcare, and energy. Although not highlighted as the panel, Brighteye also recently outlined how talent mobility technology can be seen as a missing layer in Europe’s workforce infrastructure and an investment opportunity. | TaskHer Greenworkx Jack and Jill Breakroom | June 2026 · David Guérin (Brighteye) · The Missing Layer in Europe's Workforce Infrastructure Jan 2025 · WEF · Future of Jobs Report 2025 Dec 2025 · OECD · Skills Outlook 2025: From Skills to Labour Market Opportunities |
B. New Ways to Organise Work | Tools that improve flexibility, devolve power, reduce admin, and make work more humane. This includes AI-enabled access to rights and protections that were previously gated by cost, with impact assessed against realistic alternatives. | Valla Organise | |
C. Organisational transformation tooling | AI transition infrastructure: change-management tools, workforce planning and skills forecasting, AI upskilling, and human-in-the-loop deployment systems that preserve worker voice and value. The gap is especially strong where organisations need to redesign work, not just buy software. | Mindstone Beamery Visier | |
D. Education and skills infrastructure | Education is slow-moving, but the opportunity is significant. The strongest wedges are likely to be assessment reform, AI literacy, self-directed learning, and admin reduction. Schools are particularly hard to shift, so the most investable areas are likely to be where AI can improve signalling and free teachers to focus on higher-value human interaction. | MagicSchool AI Multiverse AMBOSS Khanmigo (Khan Academy) Sana Labs | |
E. Robotics and physical AI | The obvious opportunities are in high-impact use cases such as automating unsafe tasks in manufacturing and mining, improving surgical precision and end-to-end care, reducing pesticide exposure in agriculture, and improving crop yields and resource efficiency. The less obvious opportunities are in repetitive manual work across strained sectors like hospitals, schools, the energy grid, and food systems, where labour shortages are already acute. | Intuitive Surgical Diligent Robotics (Moxi) Saga Robotics (Thorvald) |
How to evaluate responsible AI
For AI investments where labour displacement is a factor, the panel proposed a net-value test:
Quantify the anticipated impact gain at scale (resource efficiency, health outcomes, cost of access, etc.)
Quantify the anticipated labour displacement cost at scale - not just headcount, but quality of work displaced, transition costs, and geographic concentration
Is the company deploying in an area of genuine labour shortage (care, construction, agriculture) or primarily displacing desirable work?
Does the deployment model preserve worker voice? Evidence from prior tech waves shows adoption is higher and outcomes are better where it does.
Supporting frameworks are beginning to formalise this approach in practice. For example:
ImpactVC alongside Reframe Ventures and Project Liberty, with support from Zendesk has also developed our Responsible AI Due Diligence Toolkit to help VCs and LPs apply these principles in practice.
Closing view: not abundance or collapse - but reallocation
This wave of AI is not simply creating abundance or driving collapse - it is reallocating work across the economy.
That shift won’t happen all at once, or evenly. But it is already visible to those paying attention. Framing the future as either techno-utopian or catastrophic misses the more important point: the transition itself is where the real impact and opportunity lies.
For impact investors, the opportunity is to back the infrastructure that helps people and institutions navigate this transition:
On the supply side: tools that enable mobility, reskilling, and career navigation
On the demand side: companies where the net societal benefit clearly exceeds the cost of labour displacement
There are a few domains that were flagged to pay attention to:
Labour mobility & career infrastructure Sector-specific hiring platforms, skills verification, and tools that help people move into high-demand sectors (e.g. construction, healthcare, energy)
Coaching and career transitions Products that support individuals to navigate increasingly non-linear career paths, particularly as roles evolve or disappear
New ways to organise work Tools that increase flexibility, reduce administrative burden, and expand access to rights and protections
Organisational transformation tooling Workforce planning, AI upskilling, and human-in-the-loop systems that help organisations adapt while preserving worker value
Education and skills infrastructure AI literacy, new forms of assessment, and tools that improve signalling while freeing educators to focus on high-value interaction
Robotics and physical AI High-impact applications in sectors like manufacturing, healthcare, agriculture, and energy - particularly where work is unsafe, repetitive, or facing acute labour shortages
Reading list
28 May 2026 | Rt Hon Alan Milburn (Department for Work and Pensions) | Young People and Work: Interim Report
Apr 2026 | Jan Lynn-Matern, Nic Newman & Mario Baros (Emerge Capital) | Human Capital Development: Market Deep Dive
7 Apr 2026 | Hannah Slaughter (Resolution Foundation) | Labour Market Outlook Q1 2026
17 Mar 2026 | Project Liberty | The AI Jobs Crisis: Real or Fear?
8 Dec 2025 | Louise Marston (Resolution Ventures) | A Fork in the Road: How AI Can Help or Hinder the Path to Better Work
May 2024 | Eric Hazan, Anu Madgavkar, Michael Chui, Sven Smit, Dana Maor et al. (McKinsey Global Institute) | A New Future of Work: The Race to Deploy AI and Raise Skills in Europe and Beyond
n/d | Reframe Venture | Scenario II: Labour / Productivity
n/d | UNICEF Global Learning Innovation Hub | EdTech for Good: A Global Framework for Safe, Inclusive and Impactful EdTech
27 Nov 2025 | Albert Howard & Revent | Tools for the Commons
14 Oct 2025 | Sam Baker, Jan Erik Solem & Søren Halskov (Planet A Ventures) | Robots in the Real World: Mythbusting Physical AI
