Signaux actuels de l'IA
Des indicateurs en temps réel sur la vitesse d’adoption de l’IA par les entreprises britanniques, et ce qui les freine.
AI copilots in everyday work
AI copilots are turning “manual drafting and research” into assisted workflows, making writing, planning, and analysis faster and more scalable. This is reshaping how knowledge workers spend their day.
Avant
AI‑assisted work was rare in enterprises pre‑2021.
Aujourd'hui
Over 50% of enterprise knowledge workers now use an AI assistant weekly.
Impact & Échelle
Early pilots reporting around 3x efficiency gains on routine tasks.
Dola: The Smart AI Assistant
Dola is a breakthrough AI assistant that launched in Great Britain in November and reached 100% in Google Trends within just 3 months. It's now one of the top 5 apps in the UK iPhone and Android charts.
Avant
Personal AI assistants were fragmented and often lacked deep integration.
Aujourd'hui
Dola has become a top-tier choice for GB users, dominating app store charts in April 2026.
Impact & Échelle
Rapid viral growth reaching 100% Google Trends peak in 90 days.
Sky News: Outsize AI Productivity Boost
A special Sky News segment presented by Economics Editor Ed Conway highlights IMF research showing that the UK is uniquely positioned to gain an outsize productivity boost from artificial intelligence, provided the transition is managed effectively.
Avant
The UK historically struggled with a persistent productivity puzzle compared to other major G7 economies.
Aujourd'hui
An estimated 68% of jobs in the UK are exposed to AI, signaling the highest potential for AI-driven productivity acceleration among peer nations.
Impact & Échelle
68% of UK jobs exposed to AI, indicating the highest potential national productivity leap among developed economies.
a16z: AI Spend Per Employee (The Widening Gap)
Analysis by Andreessen Horowitz (a16z) of Ramp enterprise spend data shows an extraordinary divergence in AI spending. While median firms spend $12/month per employee, top 1% power users invest up to $7,500/month per employee across LLM subscriptions, coding agents, APIs, and GPU compute.
Avant
In early 2024, median enterprise AI spend was ~$2.50/mo per employee, and top 1% firms spent ~$1,000/mo on basic chat tools.
Aujourd'hui
By July 2026, Top 1% spend skyrocketed to $7,500/mo per employee, Top 10% surged to $660/mo, and Median rose to $12/mo.
Impact & Échelle
Top 1% AI-native firms invest >600x more per employee than median firms, cementing a massive compound execution and delivery velocity advantage.
What is the Value of Monthly AI Investment to a Dev Team?
Quantifying the economic return of equipping software engineering teams with paid AI tooling reveals that a tiered monthly investment unlocks exponential engineering leverage, dramatically cutting development cycles and overhead.
Inline code completion, syntax lookups & basic chat assistance (GitHub Copilot, ChatGPT Plus).
+5–10 hrs/mo saved (2x–4x ROI, ~£500 value)
Context-aware coding agents, multi-file refactoring, test generation & codebase search (Cursor Pro, Claude 3.7 Sonnet).
+25–35 hrs/mo saved (15x–20x ROI, ~£2,200 value)
End-to-end agentic workflows (Claude Code, Opus/Sonnet API swarms), autonomous PR creation, full migrations & CI bots.
+40–60+ hrs/mo saved (25x–35x+ ROI, ~£4,500 value)
Avant
Organizations restricted dev AI tools to free tiers or £20/mo single-file completion tools to avoid direct SaaS spend.
Aujourd'hui
Full-stack AI workflows and autonomous coding agents save 25–60+ engineering hours per developer every month on architecture, coding, testing, and reviews.
Impact & Échelle
15x–35x+ Monthly ROI: A £20 to £200+/mo investment returns £500 to £5,000+/month in reclaimed senior engineering capacity per seat.