Leading AI at scale: why strategy must move beyond the CIO’s office

AI has moved from novelty to necessity at extraordinary speed. Today, 62 per cent of organisations are already experimenting, embedding it into everyday work across functions.

 

Yet governance hasn’t kept pace. Nearly half of CIOs (48 per cent) remain formally responsible for AI strategy, even though 88 per cent of generative AI usage now happens outside IT.

 

That mismatch is holding businesses back. AI adoption is accelerating, but value creation isn’t. Projects proliferate, pilots multiply – and yet only a small fraction deliver measurable return. The issue isn’t ambition or technology, it’s ownership.

 

To unlock AI’s real impact, leadership needs to move closer to where the work actually happens.

 

Why centralised AI leadership is reaching its limits

 

CIOs sit at the intersection of innovation, security and scale. They have to ensure an ROI from emerging technologies while maintaining control, compliance and resilience. It’s a demanding brief – and AI has stretched it further.

 

Every request for experimentation, proof of concept or deployment shouldn’t be going through a single office. With responsibility centralised, CIOs become bottlenecks by default, no matter how capable they are. Meanwhile, finance teams, marketers, HR leaders and operations managers are already using generative AI tools independently to solve day-to-day problems.

 

The result is an “adoption-value gap”. AI is everywhere, but only around 5 per cent of initiatives actually deliver. Central control slows execution, while decentralised experimentation lacks strategic direction. Something has to change.

 

Moving AI leadership to the edge

 

The organisations seeing real gains from AI are shifting its ownership outward. Instead of expecting IT to lead every initiative, they empower department heads to define, test and scale AI use cases within their own domains.

 

This makes practical sense. Functional leaders understand their workflows intimately. They know where delays occur, where automation will help and where human judgment is essential. With the right tools and guardrails, they’re far better placed to turn AI into measurable outcomes.

 

At Freshworks, we’ve embraced this distributed model ourselves – with IT setting the standards, and teams across the business owning the execution. The results have been impressive:

• Customer support: AI agents now resolve 34 per cent of chat tickets. Human agents focus on more complex cases, boosting agent productivity by 25 per cent and cutting new-hire ramp time from six months to three.
• Engineering and quality: Developers use AI to code, while QA teams generate test cases and automate validation. Cycle times are down by up to 50 per cent, and some debugging tasks have shrunk from hours to minutes.
• Web and digital: Page creation that once took weeks now takes hours, allowing teams to prioritise higher-impact initiatives.
• IT operations: Automated ticket routing and resolution have improved response times and employee experience company-wide.
• HR and recruiting: AI-powered Slack tools accelerate CV screening and onboarding, reducing friction in hiring.

None of this has required IT to resort to micromanagement. What it did require, however, was trust, clarity and shared accountability.

 

The CIO’s role doesn’t shrink, it evolves

 

Distributed leadership doesn’t mean decentralised chaos. CIOs remain essential, but their focus shifts from ownership to orchestration.

 

Instead of driving every AI project, CIOs define the frameworks: data standards, security policies, ethical guidelines and approved tools. They ensure teams can move quickly within clear boundaries.

 

For non-technical leaders, this shift can feel intimidating. CIOs can ease the transition by prioritising intuitive AI tools, investing in AI literacy programmes and creating cross-functional “AI champion” networks to share lessons learned. Anchoring experimentation to KPIs – forecasting accuracy, time-to-hire, operational efficiency – keeps efforts grounded in business value.

 

Orchestrated leadership is how AI delivers at scale

 

AI doesn’t succeed through control or experimentation alone. It succeeds when leadership is shared: vision and governance at the centre, execution at the edge.

 

Organisations that get this right move faster, learn quicker and see returns sooner. CIOs become conductors rather than soloists, while department leaders turn AI from an abstract strategy into everyday advantage.

 

That’s how AI starts working with people – amplifying talent, accelerating outcomes and delivering value where it matters most.

---

For more information, visit freshworks.com

---

Dennis Woodside, CEO and President, Freshworks