Global Lighthouse Network 2026
Page 31 of 56 · WEF_Global_Lighthouse_Network_2026.pdf
Evolution of artificial intelligence (AI) at Lighthouses FIGURE 19
Notes: 1. ROI calculation: total investment divided by total financial benefit (to date, at time of application).
Source: Global Lighthouse Network.
Overcoming the “GenAI paradox” by focusing on high-impact
vertical applications
Nearly eight in 10 companies have deployed GenAI
in some form, yet about 80% report no significant
impact on earnings – a phenomenon McKinsey
calls the “GenAI paradox”.46 The gap stems from
over-investing in horizontal applications and under-
investing in vertical ones. Horizontal applications,
such as chatbots, improve productivity but deliver
incremental value. Vertical applications, by contrast,
are function-specific and transformative: because
they directly influence pricing, demand, risk and
asset utilization, vertical applications are more
likely to drive measurable revenue uplift, margin
expansion and cost reduction than horizontal
applications. However, vertical applications
are harder to implement, as they require deep integration with existing systems and domain-
specific data (Figure 20).
Lighthouses are breaking through by focusing
on high-impact vertical applications which, when
assetized47 as enterprise capabilities, provide a
competitive edge. For example, those competing
on customization or speed to market, such as
Haier in Shanghai, China and Eaton in Changzhou,
China, deploy GenAI for idea generation, research
curation and sales enablement. Rather than
adding operational complexity, AI streamlines it,
transforming data and design inputs into first-
time-right prototypes that compress the time from
ideation to industrialization (Figure 21).
Global Lighthouse Network: Rewiring Operations for Resilience and Impact at Scale
31
In-house
developed
(%)
Cost
($K)
Speed to
deploy
(months)
ROI
1
D
evelopment approaches in 2025
A
nalytical
AI/MLGenAI
AI agents680
370
22869%
71%
88%12
7
74.2
1.4
1.3
IAne G stnegA I A LM/IA lacitylan AAnalytical AI/ML and GenAI solutio ns, by Lighthouse cohort
% of top solut ions presen ted for Lig hthouse designat ion
16%22% 20%30% 30%32%40%42%44%55%69% 69% 71%
60% 62%10% 22%23%% 3% 8
9%
1 2 3 4 5 6 7 8 9 0 1 1 1 2 1 3 1 4 1 5 1Sep ‘18
Jan ‘19
Jul ‘19
Jan ‘20
Sep ‘20
Mar ‘21
Sep ‘21
Mar ‘22
Oct ‘22
Jan ‘23
Dec ‘23
Oct ‘24
Dec ‘24
Sep ‘25
Dec ’25
Ask AI what this page says about a topic: