AI Productivity & GDP Impact

AI inventory management: 30% profit uplift, 25% overstock reduction, sell-through rates +15-20%, forecasting accuracy 85-90%, stockouts -14%. Q3 2025 operating profit jumped 40%. Energy costs cut 14% in pilot stores.

Estimated impact: 30% profit uplift; 25% less overstock; Q3 2025 op profit +40%

AI could add 0.4-1.3 percentage points to annual labour productivity growth in G7 economies over 10 years. But actual 2024 OECD productivity growth was only 0.4%, and AI impact is 'not yet evident in the productivity statistics'.

Estimated impact: 0.4-1.3 ppt annual labour productivity growth (projected, not yet measured at macro level)

AI completes 95% of IPO prospectus in minutes (previously 2 weeks for a 6-person team). 30% median productivity gain in two specific use cases. No meaningful economy-wide AI-productivity relationship found yet.

Estimated impact: 95% of IPO prospectus automated; 30% gain in targeted use cases

Virtual Factory reduces production planning costs by up to 30%. Collision checks reduced from 4 weeks real testing to 3 days virtual simulation. First car plant with end-to-end AI-controlled surface inspection in series production.

Estimated impact: 30% planning cost reduction; collision checks 4 weeks to 3 days

AI logistics 'Control Tower' reduced inefficient truck movements 15%. Predictive AI slashed production waste up to 80% on key lines. Autonomous media-buying agents save $100M+/year in one division. Supply Chain 3.0 targets $1.5B annual savings.

Estimated impact: 15% truck waste reduction; 80% production waste cut; $100M+ media savings

Average 15.5% boost in radiograph report completion efficiency, up to 40% for some radiologists. AI breast cancer detection: 90% sensitivity vs 78% for radiologists. KMC Manipal: 20-30 additional patients served daily.

Estimated impact: 15.5% average efficiency gain; up to 40% for top performers

27% productivity improvement across AI-deployed factories (2020-2024). Waste reduction improved 41%. Kilbourn factory: OEE +16%, production waste -42%. 23,000 employees trained in AI.

Estimated impact: 27% productivity gain; 41% waste reduction across deployed sites

CEO Sundar Pichai stated 25%+ of all new Google code is AI-generated, reviewed by engineers. ~10% improvement in engineering velocity reported.

Estimated impact: 25% of new code AI-generated; 10% engineering velocity boost

Microsoft 365 Copilot: Forrester TEI study found 112-457% ROI. Internal sales org saw 9.4% revenue/seller increase, 20% higher close rates. Document collaboration +29%. Healthcare admin loads -23%. 60%+ Fortune 500 adopted.

Estimated impact: 112-457% ROI; 9.4% revenue per seller; 29% document productivity

1M+ warehouse robots. Automated centers see 25-50% throughput boost. Sequoia system stores inventory 75% faster. Packages per employee per year: 175 to 3,870 over a decade (22x). Expected $12.6B savings 2025-2027.

Estimated impact: 25-50% throughput increase; 22x packages/employee over decade; $12.6B projected savings

6,000+ developers using AI coding assistants save 1.5-2.5 hours per week. DB Lumina AI research tool (5,000 users) saves analysts 30-45 min on earnings templates, up to 2 hours on full research reports.

Estimated impact: 6,000 developers x 1.5-2.5 hrs/week = 9,000-15,000 hours saved weekly

COiN system processes 12,000 commercial credit agreements in seconds, saving 360,000 work hours annually. LLM Suite deployed to 140,000 employees. Productivity in AI-deployed areas rose from 3% to 6%.

Estimated impact: 360,000 work hours saved annually; $1.5B annual AI business value

AI chatbot deployed for customer service. Instead of cutting jobs, 8,500 call centre workers retrained into interior design advisors. Contributed to $1.4B revenue uplift. Improved demand forecasting accuracy.

Estimated impact: $1.4B revenue uplift; 8,500 workers upskilled (no layoffs)

AI assistant handles two-thirds of all customer service chats (~2.3M conversations/month), equivalent to 700 full-time agents. Resolution time dropped from 11 minutes to under 2 minutes. Cost per transaction fell 40% ($0.32 to $0.19).

Estimated impact: Equivalent to 700-800 FTE agents; $40M projected profit improvement in 2024

Controlled study: developers completed tasks 55.8% faster with GitHub Copilot (1h11m vs 2h41m). Copilot now writes ~46% of code for users with 88% acceptance rate. 15M+ developers using it.

Estimated impact: 55.8% faster task completion for developers