The First 90 Days as a Pharma Chief AI Officer: A Practical Playbook

The Role in Context: Why the CAIO Playbook Is Unsettled

The Chief AI Officer role in pharma is still settling into shape. Different pharma organizations have positioned the role differently: in some, it reports to the CEO; in others, it reports to the CIO or to a Chief Digital Officer; in others, it sits within R&D. The reporting line shapes the role’s authority and the work the CAIO can do in the first 90 days.

The role’s mandate also varies. Some organizations have hired a CAIO to drive AI-enabled drug discovery; others, to lead AI in manufacturing and supply chain; others, to coordinate enterprise-wide AI strategy and governance. The mandate determines what the CAIO should be doing on Day 1, and CAIOs who walk in without clarity on the mandate spend weeks discovering it through trial and error.

McKinsey’s coverage of how pharma is rewriting the AI playbook describes the broader industry context: pharma cannot simply drop an AI model into an existing workflow and expect transformation. The work requires a big-picture view of how AI fits into strategy, and the CAIO is typically the executive accountable for producing and executing that view. The role’s value is in the strategic synthesis, not in the technical execution.

The first 90 days are therefore about establishing the conditions for the synthesis: clarifying the mandate, mapping the current state, building the governance, selecting the initiative that demonstrates the value, and positioning the role with the board. Executive transition frameworks — Watkins’ First 90 Days, McKinsey’s CIO transition guidance — provide useful general structure, but they have to be adapted to pharma’s specific constraints.

Days 1-15: Mandate Clarification

The first 15 days should be dominated by mandate clarification. The CAIO walks in with a hiring document — a job description, a charter, a set of board-level expectations — and reality typically diverges from these documents in important ways. The mandate clarification work surfaces the divergence and produces an aligned understanding among the CAIO, the CEO, the executive committee, and the board.

Five specific conversations need to happen:

With the CEO. What outcomes is the CEO measuring the CAIO against in the first year, the second year, and the third year? What level of autonomy does the CAIO have on resource allocation? What is the budget envelope, and what is the process for adjusting it? What are the constraints the CEO knows about that are not in the hiring document?

With the CFO. What is the actual current AI spend, distributed across the organization? How is it tracked, and what visibility does Finance have into the ROI? What is the budgeting cycle, and how does AI investment fit into it?

With the CIO or CTO. What is the technology stack underlying AI work today? Who owns the infrastructure decisions? Where are the boundaries between the CAIO’s strategy authority and the CIO’s execution authority?

With the Chief Compliance Officer / General Counsel. What are the regulatory exposure considerations the CAIO needs to understand? What AI use cases are currently under regulatory scrutiny? What governance and documentation the legal function requires?

With the Chief Scientific Officer or Head of R&D. If the CAIO mandate touches drug discovery and development, the relationship with R&D leadership is foundational. What does R&D expect from the CAIO? Where are the friction points?

The output of the mandate clarification phase is a documented understanding — not a public document, but a working memo the CAIO produces and validates with the CEO — of the mandate, the constraints, the authority, and the success measures. The memo is the reference point for everything that follows.

Days 15-30: AI Inventory and Stakeholder Mapping

The second phase produces a comprehensive AI inventory and a stakeholder map. Neither is glamorous; both are foundational.

AI inventory. The CAIO needs to know what AI is actually running in the organization — production deployments, pilots, vendor-embedded features, employee-driven experimentation with generative AI tools, and shadow deployments that have not been disclosed. The inventory is invariably larger than the executive team realizes. Vendor-embedded AI features in MES, LIMS, CTMS, and similar platforms add up; employee use of public LLM tools for work tasks is often substantial; pilots scattered across business units accumulate.

The inventory has to be classified — by use case, by risk tier, by business owner, by validation status, by spend. The classification produces the portfolio view that subsequent decisions depend on. As described in Pharmaceutical Executive’s coverage of mastering AI budgets, the inability to inventory spend coherently is a primary cause of pharma AI investment dysfunction.

Stakeholder mapping. The CAIO needs to understand who is affected by AI decisions, who has formal authority, who has informal influence, and who is currently leading work that the CAIO will eventually need to coordinate or absorb. The mapping is more political than the inventory and is best done through one-on-one conversations rather than through formal stakeholder interviews.

Days 30-60: Governance Setup

By Day 30, the CAIO has the mandate clarity, the inventory, and the stakeholder map. The next 30 days establish the governance structure that will hold the AI portfolio together.

The governance structure has four components that should be set up in the first 60 days:

AI governance committee. Cross-functional, with representation from IT, Quality, Regulatory, the major business units, and Finance. Chartered to make portfolio-level decisions: which initiatives advance, which are paused, what resource allocation is approved. The committee is the CAIO’s primary instrument of organizational influence and should be designed accordingly.

Tiered AI risk classification. A documented classification framework that maps AI use cases to risk tiers, with corresponding governance expectations. Tier 1 might be low-impact internal productivity tools with minimal review; Tier 3 might be high-impact deployments affecting patients, regulatory submissions, or major financial decisions, with full governance review. The framework aligns with FDA’s risk-based posture and provides the operational language the committee uses.

Use case intake process. A documented intake process for new AI initiatives, with criteria for advancement, gates for review, and ownership clarity. The intake process prevents the proliferation of pilots that consume capacity without producing strategic value.

Performance and value tracking. A documented framework for tracking the performance and value of AI initiatives. Without this, the governance committee makes decisions on impressions rather than evidence, which produces poor portfolio outcomes.

Days 45-75: The First Signature Initiative

By Day 45, the CAIO should be identifying the first signature initiative. The signature initiative is the high-visibility, high-value work that demonstrates what the CAIO role is for. The selection is consequential: a signature initiative that succeeds establishes the role; one that fails produces lasting damage to the CAIO’s credibility.

Several criteria shape signature initiative selection:

Demonstrable value within 12 months. The initiative should produce measurable outcomes that can be communicated to the board within the first year. Long-cycle initiatives may be more strategically important but are poor signature choices because their value materializes after the credibility window closes.

Cross-functional dependency. The initiative should require coordination across functions that the CAIO uniquely is positioned to facilitate. Initiatives that any function could lead alone do not demonstrate the CAIO’s distinctive value.

Visible to the board. The initiative should produce evidence the board can recognize. Internal capability investments may be strategically critical but are poor signature choices because they are invisible to the board.

Aligned with stated strategic priorities. The initiative should advance one of the organization’s explicit strategic priorities. Initiatives that are technically interesting but strategically orthogonal produce diffuse credibility rather than concentrated.

Bounded enough to be tractable. The initiative should be scoped to be deliverable within the available capacity. Over-ambitious initiatives consume capacity without producing the demonstrable value they were selected for.

Common signature initiative categories in pharma:

• AI for clinical trial efficiency. Site selection, patient identification, protocol design support.
• AI in regulatory submissions. Document generation, response drafting, submission quality assurance.
• AI in manufacturing predictive maintenance. Equipment health prediction, downtime reduction.
• AI in pharmacovigilance. Adverse event triage, case processing acceleration.
• Generative AI for productivity. Enterprise-wide deployment with governance, training, and value tracking.

The selection depends on the organization’s specific opportunity landscape and the CAIO’s mandate. The discipline is in selecting based on the criteria, not on what is intellectually interesting or fashionable.

Days 60-90: Board-Level Positioning

By Day 60, the CAIO should be preparing for the first substantive board engagement. The engagement establishes the CAIO’s relationship with the board and sets the expectations against which future work will be evaluated.

The board briefing should cover:

The current state. What AI is running, what value it is producing, where the risks are. The board briefing should be informed by the inventory but should not present the inventory itself; the board needs synthesis, not raw data.

The strategic framing. Why AI matters for the organization, where the highest-value opportunities are, what the organizational capability gap looks like. The framing situates the CAIO’s work in the context of board-level strategic priorities.

The governance posture. What governance is being established, how risk is being managed, what regulatory exposure exists. The governance discussion gives the board confidence that AI is being managed responsibly, which is a precondition for the board’s willingness to back ambitious initiatives.

The signature initiative. The selected initiative, the rationale, the timeline, the success measures. The signature initiative gives the board something concrete to track and provides the basis for the next year’s board-level conversation.

The 12-month roadmap. What the CAIO will deliver, what investments are required, what decisions the board needs to make. The roadmap establishes the cadence of board engagement and the budget envelope for the year.

The board briefing is high-stakes. CAIOs who deliver a focused, evidence-based, strategically aligned briefing establish themselves as credible executive leaders. CAIOs who deliver a vague or overly technical briefing position themselves as functional managers rather than strategic leaders. The difference compounds over the following year.

Pitfalls to Watch For

Several pitfalls consistently derail newly appointed CAIOs. Avoiding them is more important than executing any single component of the playbook perfectly.

Pitfall 1: Becoming the AI use case approval bottleneck. CAIOs who personally approve every AI use case produce a bottleneck that frustrates the organization and consumes the CAIO’s strategic capacity. Governance should produce decision rights distributed across the governance committee, not centralized in the CAIO.

Pitfall 2: Over-investing in technical capability. CAIOs with strong technical backgrounds often over-invest in building technical capability and under-invest in governance, strategy, and organizational change. The role’s leverage is in strategic synthesis, not in technical execution.

Pitfall 3: Picking a signature initiative that fails. A failed signature initiative produces credibility damage that is hard to recover from. The selection should be conservative on tractability even at the cost of being slightly less ambitious.

Pitfall 4: Underinvesting in regulatory and quality engagement. Pharma is a regulated industry, and CAIOs who treat regulatory and quality as compliance overhead rather than as strategic partners discover that initiatives stall at regulatory review. The CCO and the head of QA should be among the CAIO’s most engaged stakeholders.

Pitfall 5: Failing to connect with the CIO. The CAIO’s strategy authority depends on the CIO’s execution capacity. CAIOs who treat the CIO relationship as competitive rather than collaborative consistently struggle to deliver. The CIO partnership is foundational and should be invested in heavily in the first 90 days.

Pitfall 6: Producing too much documentation, too little decision. The 90-day playbook produces real artifacts — the mandate memo, the inventory, the governance charter, the signature initiative plan. But the artifacts are means, not ends. CAIOs who produce comprehensive documentation without driving decisions appear to be working but are not actually leading.

The strategic implication is that the first 90 days are about establishing the conditions for sustained leadership, not about achieving everything. A CAIO who completes the playbook reasonably well is positioned to deliver real value over the following two to three years. A CAIO who over-optimizes the first 90 days at the cost of sustainable patterns produces a strong start that does not compound.

The relationship between CAIO tenure and value delivery

CAIO roles in pharma have shown high turnover in their first generation. Some of this is natural — the role is new, fit is uncertain, expectations are calibrated through experience. But some is a function of the playbook problem: CAIOs without a clear early playbook flail in ways that erode executive confidence, and the role is too high-profile to recover from extended ambiguity. CAIOs who execute the first 90 days deliberately produce confidence early and tend to have longer tenures and more cumulative impact.

Why pharma CAIO roles differ from other industries

Pharma CAIOs face constraints that CAIOs in other industries do not: GxP regulatory exposure, longer development cycles, higher individual decision consequence, and slower organizational change patterns. CAIOs transitioning from other industries should anticipate that the playbook needs adaptation. Patterns that worked in financial services, retail, or technology may not work in pharma without modification. The discipline is to learn the pharma-specific constraints rapidly and adapt accordingly.

As the Deloitte analysis on executive transitions emphasizes, the first 90 days are not actually the period when transitions succeed or fail. The patterns established in the first 90 days play out over the following 18-36 months, and CAIOs who treat the 90-day window as the test rather than as the foundation often miss the longer horizon over which the role’s value materializes.

The CAIO and the AI Center of Excellence question

A recurring strategic question in the first 90 days is whether to establish an AI Center of Excellence (CoE). The answer depends on the mandate, the organizational structure, and the existing AI investment level. CoEs work well when the CAIO needs to consolidate scattered investment, build shared capability, and produce a center of organizational expertise. CoEs work poorly when business units have strong AI capability of their own and view a CoE as imposing constraint. The CoE decision should be made deliberately in the first 90 days, not deferred to a second-year strategic review.

References & Sources