NLW, host of the "AI Daily Brief" podcast, presents a status report on enterprise AI adoption, seeking to counter the media narrative of an "AI bubble" (imminent burst) with concrete data. He draws on external reports (KPMG, Deloitte) and, above all, on an exclusive survey conducted among his audience (over 1000 companies, 3500 use cases).
The figures show an acceleration: 90% of companies plan to increase their AI budget. 2025 marks the shift of agents from theory to practice, with 42% of large enterprises reporting having agents in production (up from 11% in early 2024). However, a "bifurcation" is emerging between leaders, who deploy AI systemically to generate new revenue, and laggards, stuck in pilots focused solely on cost reduction.
NLW's ROI survey reveals that 82% of respondents already observe a positive return on investment (including 38% reporting a high ROI). The most frequent use case remains time savings (about 5 hours saved per week), which is significant on an annual scale (7 to 10 weeks of work recovered).
However, a closer analysis shows that C-Suite executives and very small companies (1-50 employees) target and achieve more "transformational" results (new capabilities, increased output) than middle managers, who are often confined to efficiency gains. A surprising insight: although "risk reduction" is rarely cited as the primary objective (3.4%), it is the category that most often generates transformational impact (25% of cases), particularly in compliance and back-office functions overwhelmed by data volume.
In conclusion, far from a bubble, the market is entering a phase of concretization where agentic automation is starting to outperform simple assistants, validating investor optimism for the coming year.