It is the end of 2025. You have heard about AI every week for 52 weeks. In leadership meetings, it shows up as strategy. In team chats it shows up as worry, curiosity, and, occasionally, eye rolling.
In 2023, leaders spoke as if AI would quickly become the engine of growth and a cure for drudge work. Many employer brand stories followed that script, promising smarter processes, more interesting roles, and faster careers. Two years on, candidates and employees are asking a simpler question: what has actually changed in the work?
Momentum isn’t always progress, especially when you always end up back where you started.
Fathom helps you escape the loop. With insight, not intuition.
We asked a simple question on behalf of our readers: were the 2023 predictions about AI and work accurate or nonsense, and what does the evidence say about 2026?
Our verdict: the foundations were right, the human systems lagged, and employer brands win or lose on how convincingly they connect AI with value, skills, and trust.
What the 2023 study actually said
A widely circulated 2023 survey from MIT Technology Review Insights, fielded June to August, captured strong executive confidence that AI would fuel growth. Respondents were senior leaders and decision makers across multiple sectors, which means their answers signalled intent at the top of the house, not just curiosity in innovation labs.
Two signals mattered most for employer branding and the future of work story.
- Budgets were set to rise. All respondents expected increased spend on data and AI modernisation in the next year, with 46% predicting a rise over 25% (MIT Technology Review Insights, 2023).
- Efficiency was the leading near term payoff. 81% expected industry efficiency gains greater than 25% within two years, and 33% expected gains above 50% (MIT Technology Review Insights, 2023).
For employer branding and talent attraction, those findings locked in a set of expectations:
- AI would be central to transformation, not a side project.
- Skills and training would have to keep up.
- Candidates would experience AI in hiring and at work, and would judge employers on how that felt.
The sections that follow treat the 2023 survey as the baseline, then run a simple test for each claim: 2023 forecast vs 2025 reality.
1) Transformation driver
2023 forecast
AI would become a primary driver of business transformation, not just a back office tool (MIT Technology Review Insights, 2023).
In 2023, that language often meant big abstract promises. Strategy decks talked about “AI powered services” and “AI at the core of the operating model”, while individual teams were still stuck in spreadsheets and manual reconciliations. The expectation was that, by 2025, AI would be visible in product roadmaps, customer journeys, and job design.
2025 reality
92% of organisations identify AI and information processing tech as a key driver of transformation (World Economic Forum, 2025).
Scorecard: direction correct, delivery uneven. AI now appears in most transformation agendas and board materials, but the quality of execution varies sharply by industry, function, and leader appetite. In some organisations, AI is genuinely shaping products, workflows, and performance conversations. In others, it remains a thin layer of automation on top of unchanged processes.
For employer branding, this matters because candidates are no longer satisfied with generic claims about “using AI”. They are looking for specifics: what this means for how decisions are made, how work flows through teams, and how their skills will age.
Implication for employer branding
Treat AI as part of your value narrative, not a feature of your tech stack. Candidates want to see how work changes and which skills compound.
Concretely, that means:
- Explaining how AI shows up in the work day for key roles. For example, “product managers use AI to test copy variations and sizing assumptions” rather than “we use AI in product”.
- Showing which business outcomes AI supports: quality, speed, insight, safety, or customer experience.
- Being honest about what is still in pilot and what is genuinely at scale.
The more clearly you can link AI initiatives to outcomes that matter for people and customers, the more credible your transformation story becomes.
2) Adoption vs enablement
2023 forecast
Talent and upskilling would be the critical bottleneck, not tooling. Unified governance would rise in importance (MIT Technology Review Insights, 2023).
Executives in 2023 already understood that simply buying platforms would not deliver value. They talked about “citizen developers”, “AI literacy”, and the need for shared standards on data and risk. The expectation was that organisations would quickly move from scattered experiments to coherent enablement and governance.
2025 reality
82% of workers say they have not received training on generative AI, and leaders cite workforce skills gaps as the top reason tech investments underperform (Deloitte, 2025 Global Human Capital Trends).
60% still say a single governance model is very important, confirming the 2023 emphasis (MIT Technology Review Insights, 2023).
Scorecard: forecast correct, pace of enablement too slow. The skills bottleneck is real, and governance has indeed become a board topic. What has lagged is practical, accessible training for everyday roles, and clear guidance on what is encouraged versus prohibited.
For many employees, “AI enablement” still means a one page policy, a webinar, and a vague instruction to “experiment responsibly”. That gap between rhetoric and reality shows up quickly in reviews and referrals.
Implication for employer branding
Publish the skills map, learning paths, and time budget. Make your governance visible and human readable.
In practice, that looks like:
- Naming the specific gen AI skills you expect people to build and how you support that learning. For example, prompt design, workflow automation, or basic data literacy.
- Stating how much time is realistically available for experimentation: “engineers have one day per month for AI exploration and skill building”.
- Turning governance into something people can actually use. Replace dense legal documents with a clear set of “do”, “do with approval”, and “do not” examples.
Candidates who care about growth, mobility, and responsible practice will read those signals closely.
Helping HR, talent acquisition, employer branding, and company culture professionals find careers worth smiling about.
3) TA scale vs candidate experience
2023 forecast
Generative AI would enter sourcing and screening at speed, improving throughput.
Two years ago, many TA teams were drowning in requisitions and manual outreach. AI looked like a lifeline: tools that could write outreach messages, triage CVs, and keep pipelines moving during hiring freezes and surges. The promise was more speed with no loss of quality.
2025 reality
87% of organisations use AI for sourcing and screening, 41% report a more impersonal candidate experience, and 67% call personalisation the new differentiator (Korn Ferry, 2025 TA Trends).
Scorecard: volume gains, mixed experience. TA teams report clear improvements in throughput and response times, but candidates often describe the process as more opaque and transactional. The distance between a personalised outreach email and a generic one generated at scale is still very visible.
Where organisations have been more deliberate, AI supports better matching and timely nudges, while recruiters keep ownership of high stakes interactions. Where they have not, AI has simply automated the feeling of being processed.
Implication for employer branding
Design human in the loop touchpoints and disclose them. Use AI to increase relevance, not to replace presence.
Three practical moves:
- Map your hiring journeys and decide where AI can safely automate and where human contact is non negotiable. For example, you may use AI to summarise CVs, but a recruiter should always deliver offers.
- Tell candidates where AI is involved. A short line such as “we use tools that help us prioritise applications based on skills, but every CV is reviewed by a recruiter” builds trust.
- Measure experience, not just time to fill. Track candidate sentiment after AI enabled stages and adjust where friction or distrust spikes.
Your candidate experience story is increasingly a technology story and a human design story at the same time.
4) Engagement and culture
2023 forecast
Efficiency gains would translate into better engagement once toil decreased.
The 2023 narrative assumed a simple chain: less admin, more time for creative and meaningful work, therefore higher engagement. AI would take on low value tasks and free managers and teams to focus on coaching, problem solving, and innovation.
2025 reality
Global engagement sits around 23%, with managers driving most variance. Tools without better leadership and culture design do not move the needle reliably (Gallup, Workplace hub).
46% of organisations still cite culture and resistance to change as barriers (World Economic Forum, 2025).
Scorecard: assumption too optimistic. In many organisations, AI has changed task mix but not necessarily improved the experience of work. Time saved has often been absorbed by more reporting, more projects, or more change initiatives. Managers remain the key differentiator, for better or worse.
Employees also report fatigue with constant technology rollouts that are not matched by clarity on priorities or recognition for effort. AI can reduce toil, but it does not design better goals, feedback, or fairness.
Implication for employer branding
Your manager standards are branding surfaces. Bake recognition, clarity, and decision rights into the story and the practice.
This is where the culture story and AI story meet:
- Define what “good management” looks like in an AI enabled environment. For instance, how managers set expectations on tool use, experimentation, and boundaries.
- Show how you help managers navigate change. Micro learning on coaching through disruption, recognition, and basic data use often has more impact than another platform.
- Bring engagement data into your narrative honestly. If global engagement is 23%, explain where you are above or below that and what you are doing about it.
Candidates will compare your promises on culture with what they see in reviews and hear from insiders. Manager behaviour is where that comparison lands.
EBN
5) Skills disruption
2023 forecast
Upskilling would be continuous and uneven, creating pressure for evidence based learning.
Leaders expected a period of rapid skill churn, with some roles changing shape almost yearly. They talked about “lifelong learning” and “skills based organisations”, but many still relied on traditional job architectures and static training catalogues.
2025 reality
63% cite skills gaps as a barrier, and projected core skill disruption reaches 86% through 2030 (World Economic Forum, 2025).
Scorecard: forecast accurate and underpowered. The scale of disruption is at least as large as predicted, and in some sectors more acute. The gap is not only in technical skills, but in analytical thinking, resilience, and working with AI systems.
Most organisations now use the language of “skills first”, but relatively few have aligned performance management, internal mobility, and pay with that model. Employees notice when skills are praised in strategy documents but not recognised in promotions.
Implication for employer branding
Shift to skills first messaging. Show the skills people will build in year one and how learning ties to mobility and reward.
To make that credible:
- Describe roles in terms of skills and outcomes, not just responsibilities. For example, “you will build skills in data storytelling and workflow automation” is more concrete than “you will collaborate with stakeholders”.
- Link learning to visible moves. Share examples where new skills led to lateral moves, promotions, or cross functional projects.
- Publish at least a simple view of your skills operating model: how you identify critical skills, how you invest in them, and how they show up in career conversations.
A clear, practical skills story is rapidly becoming a hygiene factor for attractive employer brands.
6) Data foundations
2023 forecast
Modern platforms and unified governance would separate sprinters from stumblers. Lakehouse adoption would accelerate benefits.
The 2023 report framed data architecture as the foundation for AI value. Organisations that could bring data together in a governed, accessible way would unlock better analytics and more robust AI, while others would struggle with silos and inconsistent definitions.
2025 reality
Lakehouse adoption sits near 74% among respondents, and 99% of adopters report benefits. Consolidation reduces duplication and enables trustworthy analytics (MIT Technology Review Insights, 2023).
Scorecard: structural call correct. Organisations that invested early in modern data architectures now find it easier to build AI use cases that stand up to scrutiny. Those without a coherent data story are still wrestling with basic questions such as “which headcount number is right” and “how many candidates dropped out of stage three”.
For people functions, this shows up in workforce analytics, pay equity reviews, and talent pipeline reporting. Clean, connected data makes it much easier to spot bias, track outcomes, and share transparent metrics with candidates and employees.
Implication for employer branding
Faster data means clearer hiring signals, fewer hand offs, and more transparent timelines. Translate architecture into candidate facing benefits.
Rather than talking about “lakehouses” in your careers content, talk about outcomes such as:
- “We can give you a realistic timeline for most roles because our data is in one place.”
- “Hiring managers see the same information as recruiters, which reduces duplication and confusion.”
- “We track the impact of our AI tools on speed and fairness, and we adjust based on the evidence.”
Your technology stack becomes part of your credibility story when you link it to how people are treated.
Momentum isn’t always progress, especially when you always end up back where you started.
Fathom helps you escape the loop. With insight, not intuition.
7) TA personalisation
2023 forecast
Personalisation at scale would differentiate brands and reduce friction.
The expectation in 2023 was that AI would help tailor content, roles, and journeys to individual candidates. Done well, this would cut through noise and signal that employers understood different career stages and interests.
2025 reality
Personalisation is indeed the differentiator for 67%, yet it risks feeling automated and cold if consent and human contact are not designed in (Korn Ferry, 2025 TA Trends).
Scorecard: insight correct, implementation patchy. Many candidates now receive tailored role suggestions, content, and nurture sequences. The problem is not the concept of personalisation, but the lack of consent and clarity. When people are not sure why they are seeing certain jobs, or how their data is used, personalisation feels intrusive rather than helpful.
Implication for employer branding
Treat personalisation as a system. Define a 24 hour human touch threshold for priority applicants and disclose your policy.
Useful design choices include:
- Explaining how you match candidates to roles and content, in simple language.
- Letting candidates choose the level of personalisation they want and the channels they prefer.
- Setting a clear human response commitment for high intent candidates. A simple statement such as “if you are shortlisted, a recruiter will contact you within 24 hours of your interview” can do more for your brand than any AI feature description.
Personalisation that respects agency and attention becomes a differentiator rather than a red flag.
A brand agnostic playbook for 2026 planning
Designed to work across industries and sizes, focused on employer branding, talent attraction, culture, and organisational transformation.
Think of this as a short checklist to test whether your AI story and your people reality line up.
1. Build the AI native EVP
For your top 10 roles, list 3 AI augmented workflows, the skills gained, and the career acceleration unlocked. This shifts the value case from tools to human performance (Deloitte, 2025 Global Human Capital Trends).
An AI native EVP does not start with platforms. It starts with questions such as: how does AI change what a great engineer, recruiter, or marketer can achieve here, and how fast can their skills grow as a result. That is the story candidates care about.
Make it tangible:
- For each role, write one short “day in the life” vignette that shows AI in use, and one concrete outcome it enables.
- Highlight where AI removes friction, such as summarising notes or generating first draft analysis, and where humans still own judgement.
- Avoid vague lines like “you will work with advanced AI”. Explain what that looks like in tasks, tools, and exposure.
The more specific your examples, the more believable your EVP becomes.
2. Publish a Skills Operating Model
Use a WEF style taxonomy to spotlight rising skills to 2030 and allocate time to build them. Tie to mobility and compensation. This operationalises the 86% skill disruption outlook (World Economic Forum, 2025).
A skills operating model is not a one off slide. It is a living map of:
- Which skills matter most for your strategy.
- How you assess and track them.
- How they influence hiring, performance, internal moves, and pay.
Publishing even a simplified version signals seriousness. It also gives managers and employees a shared language for career conversations. If you can show that people who build specific AI related skills see clearer paths and better rewards, your promise of “growth” has substance.
3. Personalise without dehumanising
Set consent standards for data use and a 24 hour human touch guardrail. Use AI to craft relevance, not to replace presence (Korn Ferry, 2025 TA Trends).
This starts with trust. Candidates are increasingly literate about how their data might be used. A short, plain language statement on your careers site about what you collect, why, and how you protect it is no longer optional.
Combine that with:
- Explicit consent steps, not buried checkboxes.
- Clear opt out routes for automated matching or communications.
- Guardrails for human contact, especially at later stages in the funnel.
The aim is simple. Let AI help you show people the right roles and content at the right time, while making sure that when stakes rise, people talk to people.
4. Evidence the work, not the tool
Replace “we use AI” with before after outcomes by role. Anchor your story in measured value (MIT Technology Review Insights, 2023; World Economic Forum, 2025).
Instead of listing technologies, describe what has improved:
- “Support engineers resolve customer issues faster after introducing AI assisted triage.”
- “Recruiters spend less time per requisition on manual scheduling, and use that time for candidate coaching.”
Where you do not yet have precise metrics, be honest and describe experiments in progress. Candidates can live with incomplete data. They are less forgiving of inflated claims.
Helping HR, talent acquisition, employer branding, and company culture professionals find careers worth smiling about.
5. Manager enablement is the brand
Ship micro learning on recognition and clarity. If engagement is 23%, the manager workflow is your most influential branding surface after the career site (Gallup, Workplace hub).
Managers are the interface between big AI narratives and individual experiences. They decide how tools are introduced, how workloads shift, and how success is recognised. If they are confused or unsupported, no amount of branding will compensate.
Build your manager offer around:
- Simple playbooks for running team experiments with AI.
- Guidance on talking about skills, not just job titles.
- Micro habits for recognition, boundary setting, and psychological safety during change.
Then talk about this work in your employer brand. Candidates want to know that managers are not left to figure everything out alone.
6. Responsible AI and candidate trust
Publish what you use, what you never use, how you evaluate bias, and how to request human review. Close the trust gap highlighted by culture resistance (World Economic Forum, 2025).
A short “Responsible AI in Hiring” note can answer many of the questions candidates now bring to interviews, sometimes silently:
- Which AI tools are active in your hiring process.
- Where humans always review or override automated recommendations.
- How candidates can raise concerns or request a human review if they suspect an error or bias.
This is not only a compliance issue. It is part of your reputation as an employer that takes fairness and agency seriously.
Conclusion
The 2023 foundations were sound: invest, modernise, govern. 2025 confirms AI as the engine of transformation and spotlights the human work still to do. The promises that remain unfulfilled have less to do with algorithms and more to do with skills, leadership, and honest communication.
For HR, TA, and employer brand leaders, the opportunity in 2026 is to close the gap between narrative and lived experience. That means rooting AI stories in specific workflows, outcomes, and development paths, and being open about the trade offs and experiments along the way.
As AI becomes ordinary in tools and infrastructure, the differentiator shifts to how well organisations help people adapt, grow, and trust the systems around them. The brands that win will prove value in the work, show real skill growth, and make the experience unmistakably human.
The open questions are not trivial. How fast should organisations move when skills and expectations keep shifting. How do they balance automation with good work. Those are the questions that will shape the next wave of employer brand stories.
Takeaways
1) What did 2023 get right about AI and work?
Budgets and urgency. 46% expected spend to rise over 25%, and 81% expected strong efficiency gains within two years (MIT Technology Review Insights, 2023).
That confidence has largely held. The question now is not whether to invest, but how to target spend so it improves work rather than just adding tools.
2) What is the clearest 2025 signal?
AI is the top transformation driver for 92% of organisations (World Economic Forum, 2025).
It has moved from side topic to organising principle. For talent leaders, that means AI belongs in role design, EVP, and leadership expectations, not just in IT roadmaps.
3) Why do many AI investments underperform?
Enablement. 82% of workers report no gen AI training (Deloitte, 2025 Global Human Capital Trends).
The technology is available, but the skills, time, and guidance to use it well are not. Employer brands that show realistic learning support stand out.
4) Did TA get better with AI?
Throughput improved, but 41% say the experience became more impersonal. The fix is governed personalisation with designed human touch (Korn Ferry, 2025 TA Trends).
In other words, AI can help close gaps in speed and matching, but only human contact builds trust and connection.
5) What should we publish to earn trust?
Skills maps, learning paths, and a Responsible Hiring AI note. This addresses 63% skills gap barriers and 46% culture resistance (World Economic Forum, 2025).
Being explicit about what you offer and what you are still building turns a fuzzy “growth culture” claim into something candidates can test and believe.
6) Where does engagement fit?
At the centre. Global engagement is about 23%. Manager behaviour, recognition, and clarity remain the leverage (Gallup, Workplace hub).
AI can reduce busywork, but it cannot replace good management. Employer brands that show how they support managers are more credible.
7) What is the 2026 edge?
An AI native EVP, a funded skills operating model, and a humane candidate experience that uses AI to personalise without eroding trust.
Organisations that can show this alignment in public stories and internal practice will be better placed to hire, keep, and grow the talent they need.
References
- MIT Technology Review Insights. Laying the foundation for data and AI led growth. 2023. databricks.com
- World Economic Forum. Future of Jobs Report 2025. weforum.org
- Deloitte. Global Human Capital Trends 2025. deloitte.com
- Korn Ferry. Talent Acquisition Trends 2025. kornferry.com
- Gallup. State of the Global Workplace 2025 indicators. gallup.com
