What a Chief AI Officer's First 90 Days Should Look Like

Use the first ninety days to agree what the role owns and produce evidence of useful work.

A missing budget, unclear authority to stop projects, and an outcome nobody agreed on can leave a new officer accountable for work they cannot direct. Resolve those questions early rather than assuming the title settles them.

For a commercial mandate, I would use the first ninety days to meter AI spend, set review boundaries, agree authority, and test one workflow against an outcome the company values. The strategy should guide that work and change with the evidence.

Meter spend and set review boundaries

Before opening model access, establish a spend meter, a budget limit, and a review path for consequential actions. Track what each workflow costs and what outcome it is intended to produce. AI Cost Per Customer covers the economics; human review and the audit log cover the decision and its evidence.

Use the transition period deliberately

The first quarter gives you a defined period to understand the work, agree the mandate, and test an intervention. It is a planning horizon, not a measured deadline for every AI executive.

Look at the job that came right before this one. The Chief Data Officer averaged a tenure near two and a half years, well under a CFO or a CIO, and the goodwill tended to run out around month eighteen. As Harvard Business Review put it when the Chief Data Officer was the new hot seat, the honeymoon "often ends sharply at about the 18-month period, when they are held accountable for achieving major transformational change." Michael Watkins built an entire discipline on the same idea for every executive transition: the mistakes you make in the first three months can sink the whole tenure.

The CDO tenure research is a caution about expectations, not proof that every Chief AI Officer faces an eighteen-month deadline. Use it to agree checkpoints and achievable outcomes with the company. The hiring-failure article separates that precedent from measured CAIO evidence.

Check the authority behind the title

A title does not settle who controls the budget, approves a project, or can stop one. Ask for those rights explicitly.

IBM studied more than six hundred Chief AI Officers and found 61 percent control their own AI budget. That leaves a substantial share without budget control, but it does not tell us every other decision right. The same study found 57 percent were promoted internally. Existing relationships can help, but check which authority actually comes with the new role.

Heidrick and Struggles, who place these executives for a living, put the softer version plainly. "Too often," they write, "AI adoption within an organization becomes a responsibility that no one and everyone owns." They add that a lot of these seats are just older roles rebranded, which "blurs the definition of true AI leadership."

IBM's budget-authority finding is a reason to check what authority the role actually holds. A lack of budget control does not tell us every other decision right, and promotion from inside does not prove an authority gap. Agree the reporting line, budget, and stop decisions in writing. See who should own AI and the operating scope.

The ninety days, run backwards from a number

For a commercial mandate, work backwards from the intended result. Define the outcome and baseline before choosing tools or a pilot, so the quarter has a test the company can assess.

Days 1 to 30: meter it, then follow the money. Put spend limits and review boundaries in place before expanding access. Map where AI touches revenue, margin, cost to serve, or price. Sit with the people doing the work and the CFO. By month's end, write one page defining the business outcome, baseline, and how it will be measured. Use that page to decide which workflow deserves a test.

Days 31 to 60. Agree the mandate in writing, then choose one workflow. Specify the reporting path, budget, and project decisions rather than relying on informal access to senior people:

A reporting and escalation path that can resolve budget and priority conflicts.
A budget number you control, not one you request each time.
Written authority to stop or escalate AI projects within the agreed mandate.

Ask what happens when you want to stop a project sponsored elsewhere. If you own its result but cannot stop or escalate it, resolve that conflict before accepting the mandate. The necessary authority depends on the work, not a fixed reporting-line rule.

Once the mandate is agreed, pick one workflow that touches the outcome mapped in the first month. Redesign how the work gets done, then choose the software and evaluate the result. The hiring-failure guide explains why the scope and decision rights need to fit the work.

Days 61 to 90: test the result on a number the CFO tracks. Repeat the intervention to assess whether the outcome holds. Two wins do not establish a statistical pattern. Record the baseline, observation window, volume, cost, and failures for each run: cost to serve, margin, or pipeline that closed. Take the result and its limits to the board. A working demo and a measured commercial result answer different questions.

What burns the quarter

The quarter can disappear into open model access without metering, a strategy document detached from live work, or a tool comparison that leaves the workflow unchanged. Match governance to the current data, permissions, and consequences while you test the commercial work. Evaluate a cheaper model when it can meet the task's requirements.

The day-90 scorecard

For the commercial mandate described here, I would bring three artifacts to the day-90 review: a shipped workflow with a measured outcome and its limits; a written mandate covering budget and stop or escalation decisions; and a metric the CFO has agreed to track. Use them to agree the next checkpoint, rather than treating day ninety as a universal deadline for every AI role.

Put spend and review boundaries in place, agree the mandate, then test one workflow against a defined business outcome.

This is one spoke of the Chief AI Officer guide. Start with the pillar: Chief AI Officer: the complete guide.

Frequently asked questions

What should a Chief AI Officer do in their first 90 days?

For a commercial mandate, meter spend and set review boundaries before expanding access. Agree the reporting path, budget, and stop decisions, then test a workflow against a defined baseline and business outcome. Use the result and its limits to update the plan.

What is the biggest first-quarter mistake a new Chief AI Officer makes?

Two risks are opening model access without spend controls and treating the title as the mandate. IBM found 61 percent of surveyed Chief AI Officers control their own AI budget, so budget authority is not automatic. A strategy document needs a defined intervention and measurement plan before the company can assess its commercial result.

Should a new Chief AI Officer start with governance or a commercial project?

Start the project with controls proportionate to its data, permissions, spend, and consequences. Meter usage and put a reviewer before consequential actions. Expand the governance work as the actual scope requires it; do not wait for a commercial win to address known exposure.

How does a Chief AI Officer get real authority in the first 90 days?

Agree the reporting and escalation path, budget, and project decisions in writing. IBM found 61 percent of surveyed Chief AI Officers control their own AI budget. Check what authority the specific role holds and resolve accountability gaps before accepting responsibility for the result.

What does success look like at day 90 for a Chief AI Officer?

For a commercial mandate, review a shipped workflow with a measured outcome, a written budget and stop or escalation authority, and a metric the CFO has agreed to track. Include costs, failures, and unresolved limits. A demo or strategy document alone does not establish that outcome.

How is a Chief AI Officer's first 90 days different from a normal executive transition?

AI work can incur model costs and take consequential actions from day one. Establish spend and action boundaries early, agree authority, and test an outcome. CDO transition research can inform expectations, but it does not establish a universal CAIO tenure or deadline.

Sources

Executive-transition framework and the first-three-months thesis: Michael D. Watkins, "The First 90 Days, Updated and Expanded," Harvard Business Review Press (May 2013). Chief AI Officer budget authority (61 percent control the AI budget) and internal-promotion share (57 percent): IBM Institute for Business Value with the Dubai Future Foundation and Oxford Economics, "Solving the AI ROI puzzle" (July 2025), 600-plus Chief AI Officers across 22 geographies; figures are self-reported and correlational. AI-leadership ownership and role-relabeling quotes: Heidrick & Struggles, "2025 AI, Data, and Analytics Officers Report" (published February 2026). Chief Data Officer tenure and the eighteen-month honeymoon: Tom Davenport, Randy Bean, and Josh King, "Why Do Chief Data Officers Have Such Short Tenures?", Harvard Business Review (August 2021); the roughly two-and-a-half-year average tenure is corroborated by MIT Sloan Management Review and Korn Ferry C-suite tenure data. All survey figures are self-reported by executives, not audited results.

About the author

Jeff Brokaw is a Certified Chief AI Officer and technical-commercial operator who rebuilds the commercial layer of technically complex companies and ships AI in production, not slideware. He runs a multi-model agent system in production, governing where a model's guess becomes a company commitment, one practical example of the spend and action boundaries discussed here. He ran the commercial side of two media companies tied to $850M+ in associated exits, co-founded the AI fintech FaaStrak, which went from zero to $1M ARR in nine months, and authored the go-to-market behind a $114M institutional raise that came together in under 30 days.

The First 90 Days of a Chief AI Officer