People×AI
A strategy for working symbiotically in the enterprise
Industry 4.0 automated the machine. Industry 5.0 puts people back at the centre — not to compete with AI, but to direct teams of AI agents. The right strategy is not “who does AI replace” but a redistribution of decision rights: people keep what AI cannot yet do, and AI takes the rest.
Setting objectives, establishing values and risk appetite, deciding the exceptions, building relationships and trust, and bearing ultimate responsibility before the law and shareholders.
Processing data and repetitive tasks at machine speed, running processes end to end, detecting anomalies in real time, and escalating to a person when an exception arises.
Three gaps AI has not closed — and three strategic advantages for people
Any symbiotic strategy must start with an honest acknowledgement of what AI cannot do today. These three gaps are the map for allocating roles in every function.
Worldview
AI has no value system, no business ethics, no risk appetite and no sense of an organisation’s culture. It optimises the objective function it is given — it does not know which objective is right for the business in the long run.
The whole picture
AI sees only the data and context it is given. It does not perceive the connections across departments and legal entities, competitor moves, or market signals outside the data — the things leaders sense through experience.
Business logic
The unwritten rules, the exceptions, the relationships with customers and suppliers — logic that appears in no document. AI learns the standard process, but a business runs on its exceptions too.
The symbiotic loop: people set the intent, AI executes, people judge
Set the intent
Define the objective, the constraints, the risk threshold and the criteria for success for the AI agent.
Execute & monitor
Automate the process end to end, monitor around the clock, and handle the standard cases independently.
Judge the exceptions
Intervene only when the agent escalates: an unusual case, high value, or legal risk.
Learn continuously
Every human judgement becomes training data — business logic is progressively encoded.
Allocating People × AI across each department
Each function reads against the same framework: people move up into a new role, AI agents take over the operational tasks, and the measure of success changes accordingly.
Board of management
- Holding the worldview: vision, values, risk appetite
- Strategic decisions, M&A, capital allocation
- Legal accountability and governance across legal entities
- Strategy Copilot: daily synthesis of market and competitor activity
- What-if scenario simulation across group-wide data
- Early warning on financial, operational and compliance risk
Finance & accounting
- Judging exceptions, tax optimisation, capital structure
- Relationships with banks, auditors and the tax authority
- Strategic financial analysis for the board
- Automated posting, reconciliation and three-way PO–GR–invoice matching
- Period close and consolidation across legal entities
- Cash-flow forecasting and real-time fraud detection
Manufacturing
- Continuous improvement (Kaizen) and workplace safety
- Decisions on equipment investment and technology change
- Resolving complex incidents that require shop-floor experience
- Production scheduling (APS) balancing capacity automatically
- Predictive maintenance from IoT sensor data
- Quality control with vision AI on the line
Marketing
- Brand positioning, narrative, and local cultural insight
- Major decisions on budget and brand portfolio
- Approving sensitive content and protecting the group’s reputation
- Multi-channel content production, personalised by segment
- Real-time optimisation of advertising spend
- Behavioural analysis, trend forecasting and social listening
Sales
- Building relationships and trust, negotiating high-value deals
- Understanding each customer’s unwritten business logic
- Decisions on pricing policy and exceptional discounts
- SDR agent: prospecting, scoring and nurturing leads around the clock
- Drafting quotations and contracts from templates; the CRM updates itself
- Next-best-action recommendations for every opportunity
Warehousing & logistics
- Resolving delivery chain incidents, negotiating with carriers
- Decisions on opening or closing warehouses and designing the distribution network
- Setting priorities when resources conflict
- Real-time route and load optimisation
- Demand forecasting with automatic reorder point proposals
- WMS agent: automated coordination of receipt, issue and stocktaking
Supply & procurement
- Strategic negotiation and relationships with key suppliers
- Approval under the delegation matrix, new suppliers, high value
- Make-or-buy decisions and category strategy
- Sourcing agent: consolidating demand across entities, comparing RFQs
- Compliance agent: blacklist and conflict-of-interest checks
- Automatic PO generation, delivery tracking and payment reconciliation
Human resources
- Shaping culture, developing leaders, building engagement
- Sensitive decisions: promotion, discipline, restructuring
- Redesigning work for the People × AI era
- CV screening, initial interviews and automatic scheduling
- Onboarding agent and a 24/7 internal policy assistant
- Engagement analysis, attrition risk alerts and automated C&B
Three stages: from assistant to symbiosis
Copilot
AI supports each individual: drafting, analysing, summarising. People make every decision. The objective: build AI literacy across the organisation and standardise the master data.
Colleague (supervised agents)
Agents run standard processes within each function, with people approving at the significant thresholds. Pilot in one or two departments with clear ROI (procurement, finance) and then extend.
Symbiosis (self-operating agent teams)
Agents coordinate across departments through an orchestrator; people manage by exception and concentrate on strategy, relationships and innovation.
Four capabilities every employee needs
AI literacy
Understanding what AI can and cannot do, knowing how to delegate to an agent properly, and being able to assess the output.
Managing a team of agents
Setting objectives and constraints and supervising agents — a new management skill in place of manual operation.
Data and systems thinking
Seeing across departments, reading the whole picture that AI cannot see for itself, and connecting the links.
Judgement and ethics
Deciding in the grey areas, holding the worldview and the organisation’s ultimate accountability.