McKinsey on AI paying off in mining: production and cost lessons for engineers
Reviewed by Tom Sullivan

First reported on MINING.com
30 Second Briefing
Fifteen of 19 major miners tracked by McKinsey reported third‑quarter financial gains from AI, with the consultancy estimating potential EBITDA uplifts of 10%–15% from roughly 5% higher production and 10% lower C1 cash costs. The strongest evidence is in processing, where AI-driven mill and flotation control has reached “proven in the P&L” status, with ore-blending, stockpile optimisation and shutdown planning also rated proven, while leaching, water‑and‑reagent optimisation and most mine planning, dispatch and exploration tools remain emerging or aspirational. McKinsey’s Ferran Pujol and Freeport-McMoRan’s Ravi Malladi stressed that value depends on robust data flows, condition-based maintenance and embedded site-level tech teams, not just standalone pilots.
Technical Brief
- McKinsey’s top “proven in P&L” rating is limited to AI mill and flotation control plus shutdown optimisation.
- Computer vision, ore-blending and stockpile optimisation, and predictive maintenance for fleets/plants are only “proven”, one tier lower.
- Leaching and water‑and‑reagent optimisation tools are still classed as “emerging”, with incomplete deployment or quantified benefits.
- Eight mine-planning and operations tools – including scheduling, drill‑and‑blast, dispatch and mixed‑fleet autonomy – remain emerging or aspirational.
- All three exploration applications – AI targeting, automated core logging and grade control – are still only “emerging”.
- Digital twins of mines and plants achieve “proven” status but lack disclosed P&L‑linked figures.
- Data logistics are a hard constraint: missed oil samples or unrecorded lab results nullify predictive models’ usefulness.
- McKinsey’s 70‑70‑70 rule recommends 70% in‑house tech staff, 70% builders, and 70% drawn from senior ranks.
- Governance model assigns one executive P&L ownership for AI, with COO responsible for cross‑site replication and CFO for isolating AI’s financial contribution.
Our Take
McKinsey’s separate work on commodity trading volatility, which highlights value concentrating among firms with advanced AI and tight control of physical copper and gold flows, suggests that the 10–15% EBITDA uplift cited here will likely accrue first to miners that can integrate production-side AI with trading and marketing functions.
With only 7% of energy and materials respondents having scaled AI agents in manufacturing, yet 89% using AI in at least one function, there is a wide implementation gap that gives early movers such as large copper and critical minerals producers in the Americas a temporary cost and productivity advantage before AI tools commoditise.
In our mining database, McKinsey appears most often in pieces tied to Latin American copper and critical minerals M&A and strategy, so its 70-70-70 in-house tech talent rule is likely to be most actionable for larger operators like Freeport-McMoRan and Ivanhoe Mines that already have the scale to internalise data science and software engineering teams.
Prepared by collating external sources, AI-assisted tools, and Geomechanics.io’s proprietary mining database, then reviewed for technical accuracy & edited by our geotechnical team.
Related Articles
Related Industries & Products
Mining
Geotechnical software solutions for mining operations including CMRR analysis, hydrogeological testing, and data management.
Construction
Quality control software for construction companies with material testing, batch tracking, and compliance management.
CMRR-io
Streamline coal mine roof stability assessments with our cloud-based CMRR software featuring automated calculations, multi-scenario analysis, and collaborative workflows.
QCDB-io
Comprehensive quality control database for manufacturing, tunnelling, and civil construction with UCS testing, PSD analysis, and grout mix design management.
HYDROGEO-io
Comprehensive hydrogeological testing platform for managing, analysing, and reporting on packer tests, lugeon values, and hydraulic conductivity assessments.


