Beyond the Org Chart: AI and the Transformation of Invisible Work

An increasing number of news and research articles report that AI adoption is allowing professionals to blur and extend the boundaries of their corporate roles. With the goal of understanding how work processes might be changing in an AI-forward company, we interviewed 24 product-focused individuals at a large technology firm about how AI has impacted their own work, their work within their product team, and their professional interactions. Our conversations suggest that AI is not only changing formal role responsibilities and collaborations between those roles, but also changing informal cultural practices like professional mentoring that are key to helping professionals settle in their positions, stay engaged with their work, and grow their careers. Some of these changes are positive, such as smoother collaboration between peers, but other changes are more nuanced and put the typical career growth opportunities, like receiving feedback from professional networks and promoting leadership and mentorship, at risk. We propose steps that AI companies can take to make the invisible work more visible. Additionally, we propose efforts that individuals and leaders can take to support their colleagues through AI transformation while preserving healthy company cultures that support diverse thinking, collaboration, and informal interactions.

Summary

Main Finding

Generative AI is reshaping not only visible task output but also the “invisible work” that sustains organizational learning, mentoring, and social support. In interviews with 24 product-focused professionals at a large AI-first technology firm, the authors find that AI both strengthens and undermines critical informal practices: it smooths cross-discipline communication and accelerates individual productivity, while simultaneously blurring role boundaries, reducing human-to-human feedback and mentorship, and risking erosion of the social capital that supports career growth, well-being, and organizational learning.

Key Points

Seven dimensions of invisible work are affected by AI:
Cross-discipline communication — AI creates new boundary objects (high-fidelity prototypes, chat-based explanations) that make handoffs and mutual understanding easier.
Role clarity & division of labor — AI broadens skillsets and blurs role boundaries (e.g., designers doing “vibe coding”), which can empower workers but also cause anxiety about expectations and career paths.
Project situational awareness — AI helps generate artifacts and summaries, but there is overhead to revise and share these artifacts so others stay informed.
Reputation & accountability — People who rely on AI risk social-evaluation penalties; readers may not hold AI-generated authors accountable, creating ambiguity in responsibility.
Personal growth & learning — AI enables on-demand skill acquisition and scaffolding, especially for less-experienced workers.
Instructional support & feedback — AI supplies fast informational and task support (drafts, debugging, explanations), substituting some forms of human feedback.
Task support & collaboration — AI can act as a proxy collaborator, speeding work but reducing peer interactions that are routes to mentorship and networking.
Net effect is mixed: increased independence, speed, and new capabilities vs. decreased human social support, potential loneliness, reduced opportunities for informal mentoring, technostress, and uncertain effects on long-term career progression.
Workers reported no immediate job loss fears, but many were uncertain about long-run role evolution (5–10 years).
Gender/minority implications are a concern because invisible work has historically been borne disproportionately by these groups; AI-driven changes could reshuffle who does and benefits from invisible work.

Data & Methods

Qualitative study based on semi-structured interviews (1 hour each) with 24 professionals at a single large AI-first technology company (March–April 2026).
Participant makeup: designers (7), user researchers (7), applied/data scientists (10); three managers and the rest individual contributors; 22 of 24 agreed to automated transcription.
Recruitment: invited via manager distribution lists to sociotechnical professions; participants had to be employed at least 6 months.
Interview protocol: described regular workflows, AI use, pre-AI workflows, accuracy and trust in AI, team impacts, role changes, and job security perception.
Analysis: transcripts anonymized and coded for themes; secondary review identified recurring patterns and cross-theme correlations.
Limitations (acknowledged/inferable): small, non-representative sample from one company; self-selection and limited demographic data; qualitative design—no causal estimation or generalizable effect sizes.

Implications for AI Economics

Measurement of productivity gains should incorporate invisible-work dynamics. Short-term task-time savings from AI can be offset by longer-term costs if mentorship, knowledge diffusion, and social capital decline.
Human capital formation: AI as on-demand tutor boosts immediate skill acquisition, but reduced peer feedback and mentorship could weaken cumulative skill development and career mobility—affecting wage growth and internal promotion pipelines.
Organizational capital and innovation: erosion of informal networks and cross-disciplinary learning may slow organizational learning, reduce idea diversity, and lower innovation returns—an economic externality not captured by simple output metrics.
Labor market heterogeneity and inequality: because invisible work is unevenly distributed (often more on women and minorities), AI-driven shifts could alter who performs and who is rewarded for invisible tasks, potentially amplifying or attenuating existing inequalities.
Turnover and retention costs: increased loneliness, technostress, and diminished career signals can raise quitting risk; firms must weigh immediate efficiency gains against turnover and rehiring/search costs.
Policy and firm-level responses with economic consequences:
- Firms should develop metrics for invisible work (mentorship, feedback interactions, knowledge-sharing) and include them in performance/reward systems to avoid underinvestment.
- Design AI tools to surface provenance, encourage human-to-human handoffs, and log collaborative interactions to preserve accountability and social learning.
- Invest in structured mentoring and onboarding programs as complements to AI scaffolding to sustain long-term human capital.
- Monitor heterogenous impacts across roles, seniority, gender, and demographic groups to avoid unintended distributional effects.
Macro forecasts: aggregate productivity gains from AI may be overstated if models ignore erosion of relational capital; long-run growth estimates should incorporate potential reductions in innovation diffusion and slower human capital accumulation unless organizations actively counterbalance those effects.

Overall, the paper argues that evaluating AI’s economic impact requires going beyond minutes-per-task and output measures to include the value of invisible work: mentoring, feedback, reputation, and networks. Firms and economists should track and design for those dimensions to realize sustainable productivity and equitable outcomes.

Assessment

Paper Typedescriptive Evidence Strengthlow — Findings are based on qualitative interviews (n=24) at a single large technology firm and rely on self-reported perceptions rather than quantitative measures or causal identification, limiting strength for general claims about economic impacts. Methods Rigormedium — The study uses in-depth interviews appropriate for exploratory work and surfaces nuanced mechanisms (role boundaries, mentoring, informal practices), but rigor is constrained by small, single-firm sampling, potential selection and reporting biases, and limited information about sampling frame or analytic procedures. Sample24 product-focused individuals at a single large technology firm were interviewed about how AI affected their personal work, team interactions, and professional relationships; data are qualitative (interview transcripts) collected to explore perceived changes in roles, collaboration, and mentoring. Themeshuman_ai_collab org_design skills_training adoption GeneralizabilitySingle large technology firm — findings may reflect firm-specific culture and resources, Small sample size (n=24) limits representativeness, Possible self-selection and response biases among interviewees, Unclear coverage of roles, seniority levels, geographic locations, and demographic diversity, Qualitative, perception-based data — not directly generalizable to measurable productivity, wages, or labor-market outcomes, May not generalize to non-tech sectors or smaller firms with different HR processes

Claims (8)

Claim	Direction	Confidence	Outcome	Details
We interviewed 24 product-focused individuals at a large technology firm about how AI has impacted their own work, their work within their product team, and their professional interactions. Other	null_result	high	description of sample and data collection	n=24 0.18
AI adoption is allowing professionals to blur and extend the boundaries of their corporate roles. Task Allocation	mixed	high	changes to role boundaries / role responsibilities	n=24 0.18
AI is changing formal role responsibilities and collaborations between those roles. Task Allocation	mixed	high	formal role responsibilities and inter-role collaboration	n=24 0.18
AI is changing informal cultural practices like professional mentoring that are key to helping professionals settle in their positions, stay engaged with their work, and grow their careers. Skill Acquisition	mixed	high	informal mentoring / onboarding / career development practices	n=24 0.18
Some of these changes are positive, such as smoother collaboration between peers. Team Performance	positive	high	peer collaboration / team coordination	n=24 0.18
Other changes are more nuanced and put the typical career growth opportunities, like receiving feedback from professional networks and promoting leadership and mentorship, at risk. Skill Acquisition	negative	high	access to feedback, leadership development, mentorship (career growth opportunities)	n=24 0.18
We propose steps that AI companies can take to make the invisible work more visible. Organizational Efficiency	positive	high	organizational practices to surface invisible work	0.03
We propose efforts that individuals and leaders can take to support their colleagues through AI transformation while preserving healthy company cultures that support diverse thinking, collaboration, and informal interactions. Organizational Efficiency	positive	high	leadership and individual practices to preserve culture during AI adoption	0.03

AI tools at a large tech firm are blurring job boundaries and smoothing peer collaboration, but they are also fraying informal mentorship and feedback systems that sustain career growth.