The debate between shareholder primacy and stakeholder theory concerns the legal obligations of corporate directors. Introduction A corporation is never just a machine for making money. It is a legal institution, a social system, and a vehicle for allocating power, risk, and responsibility. That is why one of the most important debates in corporate governance continues to divide scholars, executives, regulators, and investors: should corporations exist primarily to maximize shareholder value, or should they balance the interests of all the people and institutions affected by their actions? This debate is commonly framed as shareholder primacy vs. stakeholder theory. Shareholder primacy says the corporation's central obligation is to its owners, the shareholders. Stakeholder theory says the corporation owes duties to a wider circle of affected parties, including employees, customers, suppliers, creditors, communities, and sometimes society at large. This is not a purely academ...
This post covers the shift in business management toward antifragile and AI-native operating models. It sets out the polycrisis framing, the four pillars of antifragile design, agentic AI as a structural change, and the governance that makes both work, tested against Microsoft and Safaricom.
By Kateule Sydney — Researcher and Writer | Last Verified: 2026-09-29 | Published by Kat-Syd Resources Hub
Business management in 2026: antifragile design and AI-native operations
Table of Contents
The Polycrisis Has Ended the Linear Business Model
The linear business model assumed predictable inputs, gradual change, and risks that could be managed one at a time. That assumption is now obsolete. CIMA, in its 2026 guidance for finance and business leaders, uses the term “polycrisis” to describe overlapping and interconnected global risks: geopolitical instability, artificial intelligence, climate change, and supply chain fragility, all interacting with each other at the same time.
I want you to notice what the polycrisis framing changes. It is not the volume of risk that is new. It is the interconnection. A climate event disrupts a supply chain, which raises input costs, which strains cash flow, which forces a financing decision, which triggers a governance question. Managing each risk in isolation does not work, because the shocks arrive together, not in sequence.
The three failures of the linear model under polycrisis are:
- Siloed risk management — each function manages its own risks and no one manages the interactions
- Forecast-based planning — plans built on a single forecast break the moment the environment shifts
- Static organizational design — structures optimized for a stable environment cannot reconfigure in time
The interpretive insight is that linear management was not wrong. It was fitted to a world that no longer exists. Firms that keep operating on the linear model do not fail from a single shock; they fail because the shocks compound faster than they can respond.
Antifragility Is Not Resilience, and the Difference Matters
Nassim Nicholas Taleb, in Antifragile: Things That Gain from Disorder (2012), drew a distinction that business management has only recently begun to absorb. A fragile system breaks under stress. A robust system resists stress. A resilient system recovers from stress. An antifragile system gets stronger from stress. Most firms are proud to be resilient, and they are right to be, but resilience is not the ceiling of what is achievable.
Key point: resilience is the ability to bounce back to the same state after a shock. Antifragility is the ability to end up in a better state because of the shock. The two are not the same, and firms that confuse them aim too low.
The four design pillars of an antifragile organization are:
- Optionality — hold multiple paths forward so that any single failure is survivable and some failures open new choices
- Modularity — design systems so that a failure in one part does not cascade into the whole
- Slack — keep some capacity unallocated so the firm can move when an opportunity or threat appears
- Distributed sensing — push information gathering to the edges of the organization so weak signals arrive before they become crises
The interpretive insight is that antifragility is a design property, not an attitude. It is built into the structure of the firm through specific choices about how the firm holds resources, how it organizes work, and how it hears about change.
Agentic AI Is a Structural Change, Not a Tooling Upgrade
Agentic AI refers to autonomous AI systems that can plan, decide, and execute multi-step tasks with limited human intervention. Unlike earlier AI systems that responded to prompts or produced outputs for humans to act on, agentic systems act directly within defined parameters. Gartner’s 2026 predictions treat agentic AI as a structural force that will reshape how firms operate, not just a new tool to add to the stack.
I will put it this way: adding AI to an existing process is a tooling upgrade. Redesigning the process around what AI can now do is a structural change. The firms that treat agentic AI as the first option are the firms that will extract its value. The firms that treat it as the second option will keep adding tools to processes that should no longer exist.
The three practical shifts agentic AI forces on management are:
- Process redesign, not process augmentation — assume AI does the work first, then decide what humans should keep
- Human in the loop for oversight — not for every action, but for the decisions that carry material risk
- AI-native governance — rules for what AI can decide, when it must escalate, and how its outputs are audited
The interpretive insight is that agentic AI collapses the distance between decision and action. That is powerful, and it is dangerous. The firms that get value from it are the firms that design the governance before they turn the systems on.
Governance Is the New Competitive Discipline
Governance has traditionally been viewed as a cost of doing business, a set of rules imposed from outside. In the polycrisis environment and the agentic AI era, governance is becoming a source of competitive advantage. Firms that can act quickly within clear rules move faster than firms that have to convene a committee every time a decision arises.
The shift is visible in three areas. First, AI governance: rules about what models can be deployed, what data they can access, and how their decisions are documented. Second, macro-stewardship: the recognition that sustainability, supply chain resilience, and community impact are core business model issues, not compliance overhead. Third, decision rights: explicit statements of which decisions belong at which level, and which can be delegated to autonomous systems.
The three governance disciplines that separate leading firms are:
- Explicit AI policy — documented rules for deployment, escalation, and audit of AI-driven decisions
- Integrated risk oversight — risk managed at the interaction level, not only at the functional level
- Macro-stewardship accounting — sustainability and stakeholder impact measured alongside financial performance
The interpretive insight is that governance has moved from a defensive posture to an offensive one. Firms with strong governance can act faster, because the boundaries of acceptable action are already defined.
Sustainability Has Moved from Compliance to Core Business Model
For the past two decades, sustainability was largely treated as a compliance exercise: report the numbers, meet the regulation, avoid the reputational damage. The 2026 guidance from CIMA and other professional bodies marks a shift. Sustainability is now framed as a core business model issue, with direct implications for cost of capital, supply chain resilience, customer preference, and long-term value.
I will not pretend that all firms have made this shift. Many still treat sustainability as an annual report chapter. But the firms that are moving early are discovering that sustainability is not a constraint on strategy; it is a source of strategic options. Firms that redesign products for lower material intensity reduce input costs. Firms that build resilient supply chains reduce exposure to climate and geopolitical shocks. Firms that treat employees and communities as stakeholders build reputational capital that protects them in difficult periods.
The three practical moves firms are making are:
- Integrate sustainability metrics into capital allocation — projects are evaluated on environmental and social impact, not only financial return
- Redesign for material and energy efficiency — treat resource use as a cost lever, not only a compliance obligation
- Treat stakeholders as long-term counterparties — employees, communities, and suppliers are managed for durability, not extraction
The interpretive insight is that sustainability is no longer separate from strategy. It is one of the inputs to strategy, and the firms that treat it that way build advantages the firms that treat it as reporting cannot match.
Judgement
The case evidence from Microsoft and Safaricom establishes that the shift to antifragile and AI-native operating models is not a future scenario. It is already happening at the firms that lead their markets. Microsoft has restructured around AI-native products and internal operating practices, with governance frameworks for responsible AI deployment built into product and process decisions. Safaricom operates a regulated national network where reliability is non-negotiable and continuous adaptation is the operating norm, and its approach to resilience includes the M-PESA platform, training pipelines, and stakeholder commitments that span well beyond the quarterly cycle.
My position is that antifragility is the correct frame for the current environment, and I depart from the resilience orthodoxy on this point. Resilience aims at returning to the prior state after a shock. That was adequate when shocks were occasional and recoverable. In a polycrisis environment, the firm that returns to the prior state has learned nothing and changed nothing, and the next shock finds it in the same position. Antifragility requires the firm to change as a result of the shock, and to be designed in advance so that change is possible.
The standard treatment of business management undersells three things: the interconnection of risks, the difference between resilience and antifragility, and the fact that agentic AI is a structural change rather than a tooling upgrade. Firms that treat these as three separate agenda items will continue to optimize each one in isolation. Firms that treat them as one agenda build operating models that can absorb shocks, use AI natively, and govern both with discipline.
My recommendations follow from that conclusion:
- Design for antifragility, not only resilience — build optionality, modularity, slack, and distributed sensing into the operating model so that shocks produce learning and change
- Treat agentic AI as a process redesign trigger — assume AI does the work first, then decide what humans should retain, and design the process around that assumption
- Invest in AI governance before scaling deployment — define what AI can decide, when it must escalate, and how decisions are audited, then scale within those rules
- Move sustainability into capital allocation — evaluate projects on environmental and social impact alongside financial return, and treat resource efficiency as a cost lever
- Measure risk at the interaction level — manage the connections between risks, not only each risk in isolation, because the polycrisis arrives as combinations
Business management in the polycrisis era is not about surviving shocks. It is about being designed so that shocks make the firm stronger, and the firms that build that design will be the ones still standing when the next combination of risks arrives.
References
- CIMA guidance on polycrisis and macro-stewardship — CIMA
- Gartner 2026 predictions on agentic AI and business operations — Gartner
- Antifragile: Things That Gain from Disorder — Nassim Nicholas Taleb, Random House
- Responsible AI governance and deployment — Microsoft
- Safaricom company profile and operations — Safaricom
Read Also on Kat-Syd Resources Hub
Management Principles Series: Planning, Organizing, Staffing, Directing & Controlling — Foundational management functions that govern how firms plan, organise, and control.
Functional Management — The Pillars — How core business functions including HR and operations are organised and measured.
Agency and Employment Law Series — The legal foundations of employment relationships and agency.
Written by Kateule Sydney — Researcher and Writer
© 2026 Kat-Syd Resources Hub. All rights reserved.
Kat-Syd Resources Hub — Educational case studies and analytical reference