AI is changing what’s possible, in every category, for every organization.

The bigger question is not whether to use it, but: what kind of organization do you want to become?

The majority of USA and UK firms – 78% and 71%, respectively – now use at least one type of AI technology in their work (NBER, February 2026).

Many of those are running pilot schemes, trying to uncover the theoretical possibilities without trialing them in live scenarios. The biggest risk isn’t making the wrong choice, it’s waiting too long to take action. If leaders focus only on getting AI ‘right,’ the business can drift into a direction no one consciously chooses.

It can be valuable to define the type of organization you are today, then decide the type you want to become. Leaders can generally identify their organization from our shortlist of models:

1. The tech-led organization

Tech-led organizations are where AI and automation drive speed, scale, and consistency. People set the direction, oversee, and apply judgement where it counts.

Amazon has been tech-led from the start. Years of investment, scale, and revenue growth have given it room to experiment. This approach gives Amazon space to test, fail, learn, and keep going. In his 2025 letter to shareholders, Amazon CEO Andy Jassy shed light on how the company has used AI. “You have to have people and a culture that are comfortable operating with ambiguity as you sort through the new normal,” he writes. “You need to invent and experiment like crazy. Many of these experiments will fail, and it might feel like you’re getting nowhere. But, your culture must possess the tenacity to keep at it.”

In a December 2024 DealBook interview with Andrew Ross Sorkin, Bezos said Amazon was working on roughly a thousand internal AI applications and that modern AI “can be used to improve everything” and “will be in everything.” - comparing it to electricity and echoing Andrew Ng.

2. People, tech-augmented organization

In this scenario, people lead while AI improves insight, judgment, effectiveness, and time spent on high-value work.

The Mayo Clinic is a great example of an organization that is using AI to improve care by accelerating research, supporting faster diagnosis, and giving clinicians better tools to treat patients. Through its own models, partnerships with organizations like Microsoft, and more than 200 AI initiatives, Mayo is applying AI to areas such as earlier disease detection and clinical trials. One example is StateViewer®, which interprets brain scans faster and more accurately while keeping clinicians in control of final decisions.

3. The people-centric organization

Outdoor clothing brand Patagonia’s core model is anchored in purpose, product durability and customer relationships. Still, technology contributes to its goals. Patagonia uses AI selectively to support its efforts – such as its Worn Wear resale platform, detailed in the brand’s 2025 Impact Report, which allows customers to trade in or repair their Patagonia clothing, or buy others’ second-hand items.

The brand has taken a careful, considered approach to technology, integrating it in ways that preserve its integrity and identity as an organization deeply rooted in environmental responsibility. As Head of Technology Jon Chong describes it, the goal is “high tech to touch”: using technology to strengthen human connection, not replace it.

4. The legacy organization

Technology can be used to selectively support those businesses where people lead through expertise and relationships.

Many regional banks are using AI mostly as a controlled efficiency and risk-management lever, not as a full transformation of the banking model (S&P Global, June 2026). In its 2024 Small Business Lending Survey Report, the US’s Federal Deposit Insurance Corporation (FDIC) reported how having a physical, bricks-and-mortar presence, as well as carrying on-site visits, remain integral to regional banks’ success.

So how should leaders choose?

The question is not: “How much AI can we deploy,” it’s “Where should AI create an advantage?”

Different brands, markets, and functions may need different levels of AI and automation. And this may change over time.

You need to make three choices:

• Where AI should lead – high-volume, repeatable, data-rich workflows where speed, scale and consistency matter.
• Where humans must lead – moments involving trust, judgement, ethics, creativity, relationships or high-consequence decisions.
• Where human-AI collaboration creates advantage –where AI can handle the heavy lifting, but people frame the problem, challenge the output and own the outcome.

At Oxford, we help organizations build the capability and lead the change that makes the right level of AI integration stick.

We can help you become the organization you want to be.