Wharton AI Adoption Report - 2025

Wharton published the third edition of their AI Adoption report this week, and its packed with 90 slides of valuable data for any senior leader.

The authors surveyed 800 respondents, all in senior decision making roles across HR, Legal, IT, Marketing/Sales, Operations, Product/Engineering, Purchasing/Procurement, Finance/Accounting and General Management.

All respondents were from U.S.-based enterprise commercial organizations with at least 1,000 employees and over $15M in revenue.

The goal of the survey was to get a sense check from enterprise leaders on their adoption of Generative AI, building on the results of the previous two studies in 2023 and 2024.

There are 90 slides in the report, and I encourage you to download and take the time to read in detail, but we'll summarize the key lessons we take from it here.

Gen AI moves from dabbling to daily productivity

In previous reports, the sentiment was around experimentation and proof of concept.

In this year's report, usage has evolved into a daily productivity driver with respondents across different teams using Gen-AI tools daily - 46% across the entire survey up 17 percentage points from last year.

Yet it's not consistent across each team.

You'll see here that IT is surging ahead driven by the growth in AI-assisted coding.

An interesting call out for me is that marketing and sales lag behind other departments when they are well-provided by AI modules in many of the tools that they use.

Half of top ten Gen-AI use cases directly boost employee productivity

When provided with access to the infinite knowledge machine, the big challenge for companies and their employees is to know what to do with it.

This is a challenge that companies continue to struggle with.

This chart shows the top ten use cases listed by respondents, and you can see that a good proportion of them are automating, summarizing, and simplifying the creation of content.

• Help me summarize this email.
• Help me curate these meeting notes.
• Help me draft a proposal.

The missed opportunity here is that none of these use cases are building on GenAI's coaching and critical thinking capabilities as a mentor and coworker.

Meaning there is a huge opportunity for companies to teach their people how to move GenAI from single-player mode into multi-player mode and being a valued part of a team.

Productivity might be your first step in getting value from Gen AI, but improved ideas and quality of output is the destination.

ChatGPT and Co-pilot dominate usage

Remembering that this report surveyed senior leaders in companies with more than 1,000 employees, it's interesting to see the spread of AI tools in usage.

Whilst the majority of large enterprises have deployed their enterprise productivity suites platform, (Microsoft customers have rolled out Copilot, and Google customers have rolled out Gemini as part of their Google Workspace plan), it's interesting to see that ChatGPT still dominates with 67% of companies having ChatGPT usage in their organisations.

I specifically point you to the fifth bar, which is a custom chatbot built specifically by or for your organization.

We see in larger organizations that IT have built a custom wrapper around one of the leading providers, APIs, and called it something like Acme GPT.

Whilst this may have been a pioneering strategy two years ago to provide AI safely to employees, increasingly it's becoming a hindrance as internal IT teams need to keep up with the application layer that is being built on the top of tools like ChatGPT.

Internal IT teams need to be aware that ChatGPT already exists on every employee's phone, and they know how to use it and can. If they feel they'll get the answer they need in the format they want.

Every company that we roll out to ChatGPT Enterprise to has previously rolled out Copilot, Gemini or an internal tool.

There's something to learn in that.

Tech and internal R&D take priority in Gen AI spending

This next chart breaks down how organizations of different sizes are allocating their AI budget, and it's interesting to see that it is technology tools and systems that are taking up the largest share, whether that is buying new tech or adding budget to existing technology.

Gen AI success comes from employees, from the CEO down to the intern knowing how to use this hugely powerful technology.

See that employee training ranks far lower down in terms of budget allocation across these organisations.

Three-quarters of enterprises report positive return on investment

Three-quarters of the respondents reported positive return on investment in their Gen AI programs. This chart shows the split by size of organization from $50M revenues in the second column up to those that are Tier 1 with $2B and above in the right-hand column.

Across all company sizes, just 15% of companies reported either negative ROI or too early to test, which provides convincing validation that all companies should be leaning in to their Gen-AI spending.

Training expectations for GenAI fluency remain unclear

In this final chart, it's clear that across different teams, there are varied expectations of what level of training is required to enable employees to capitalize on this new technology.

Across all respondents, 38% said they'll need extensive required training or investment.

Take a look at product development and engineering where 29% believe they will need entirely new talent.

Where will these new employees come from?

Go deeper

These slides are just a snapshot of the full report, which goes into detail across different industries, company sizes, roles, and provides commentary and guidance on how to interpret this data in the context of your own organization.

Below you will also find a prompt that you can copy into your own AI tool, then upload the full 90-page report and it will guide you through a discussion of how to interpret this report in the context of your own organization.

-------

You are an AI strategy advisor helping me interpret the 2025 Wharton–GBK AI Adoption Report in the context of my own organisation.

First, ask me 4–5 questions to understand my situation, such as:

- What industry is your organisation in?
- Roughly how large is your organisation (headcount or revenue)?
- What’s your current level of AI adoption (experimentation, pilots, embedded use cases)?
- Who is driving AI adoption internally (e.g. IT, innovation, business units)?
- What are your main goals for AI adoption (efficiency, innovation, competitive advantage, etc.)?

Once I’ve answered, summarise my context back to me, then provide a short overview of how the Wharton–GBK report is structured:

- Executive summary – headline findings on adoption maturity and leadership sentiment
- Adoption maturity stages – how organisations are progressing from exploration to scale
- Industry comparisons – which sectors are leading or lagging
- Leadership and governance – how AI ownership, skills, and risk are being managed
- Business impact and ROI – where measurable value is emerging
- Barriers and enablers – what’s holding organisations back, and what’s helping them move forward

Ask me: “Would you like to explore any of these areas in more depth?”

Offer options such as:
- Compare your industry’s position with the benchmarks
- Examine leadership and governance lessons
- Identify key use cases for our industry
- Look at case studies or metrics for ROI

Once I choose an area, draw insights from that part of the report and relate them directly to my organisation’s context and goals.

Then:

1. Summarise the key insights from the report that are *most relevant* to my type of organisation.
2. Highlight where my organisation seems to align or differ from the report’s benchmark data.
3. Suggest 3–5 tailored actions my organisation could take to advance AI adoption responsibly and effectively.
4. Recommend how to measure progress over the next 6–12 months.

Provide references to your findings.

Finally, provide a concise summary I could use in an internal presentation to my leadership team.