Executive Summary
Dawn’s GenAI Adoption Study of 40 billion-dollar revenue enterprises finds GenAI moving from proof-of-concept to production: 78% of companies are formally deploying it today, and a majority expect the technology embedded in every routine task by 2027.
Momentum is accelerating. More than half of respondents aim for enterprise-wide rollout within 6 months, and executives rate GenAI a top-three agenda item for both internal productivity (weighted score of 8.0/10) and external products (weighted score of 7.6/10). These firms view GenAI as the lever to compress cycle times, cut cost, and boost revenue simultaneously – ambition that underpins the rapid expansion of pilots into core workflows.
Usage breadth and early economics are compelling. Over 70% already run ten or more GenAI initiatives, spanning workflow automation, knowledge ops, sales enablement, and customer support. While half of “fastest” projects still take 8+ weeks from idea to production, the balance of ROI is net-positive (+25 pp) and tilts first toward cost reduction and productivity gains, with revenue growth close behind. Budgets are following the momentum: 43% now ring-fence dedicated GenAI funding, and nearly half allocate 2-5% of total IT spend to these projects.
Scaling remains messy. Security dominates anxieties – rated 6.3/10 on average as the most predominant business concern – with application-level security the single largest technical barrier (9.4/10 on average). Talent shortages, data quality gaps, and regulatory friction follow. Tooling is fragmented: a majority juggle 4+ LLMs and 4+ GenAI tools, even as OpenAI soaks up c. 60% of spend. This patchwork amplifies governance complexity and inflates integration effort.
Governance is racing to catch up. 63% now have a formal AI policy, typically owned by the CTO or engineering lead, and every firm monitors shadow-AI usage to curb risk. Firms cite easy-to-use, secure tools, workforce training, and clearer regulation as the top enablers for scale. Only 23% have cancelled GenAI projects, underscoring durable conviction despite hurdles.
In all, the findings paint a world in transition. Enterprises that pair bold deployment with disciplined security, unified tooling, and upskilled talent will widen the productivity gap; laggards risk obsolescence as GenAI-first efficiency becomes table stakes.
We hope you enjoy reading the findings as much as we did.
Profiles
Sample N=40
75% of respondents were in the US, and 25% in Europe.
We surveyed enterprise leaders across strategic and technical functions.
We recruited leaders from the “early adopting” enterprise verticals – financial services, telcos, and software service providers. All had >$1b in revenue, 60% were >$10b.
AI strategy
Journey
>75% of respondents are formally deploying GenAI, beyond experimentation and pilots
Q: Where is your company today on its GenAI‐adoption journey?
Over half of companies plan to have deployed GenAI enterprise-wide in next six months.
Q: If your company is not deploying GenAI enterprise-wide, when do you expect to reach that stage?
A majority of respondents expect GenAI to be in every routine task and daily workflow in next two years.

To integrate with older mainframe applications and customer support services… and enable low / no code development of functionality.

I want all employees to be leveraging Gen AI on a regular basis to do their daily activities.

I want GenAI to automate routine tasks like report generation and data analysis, freeing teams to focus on strategy.
Priority
While both internal and external GenAI initiatives are seen as priorities, enterprises are more focused on embedding GenAI into internal operations. Only 7 out of 40 respondents rated internal use cases ≤6 in priority, compared to 14 for external.
Q: On a scale of 0 to 10, how much of a priority is embedding GenAI into your internal operating systems (e.g., productivity tools, workflows)?

We are very bullish on GenAI and plan to have many use cases related to automation, documentation summarization, call centre Q&A response.

We are completely embracing Gen AI in all of our processes because we believe not doing so will put the company at a severe disadvantage.

It is critical for us to integrate AI into our customer-facing applications and services to increase value for the customer and improve UX.
Most enterprises want to reduce headcount, eliminate repetitive tasks, and improve their product.
Q: Which KPIs would you most like to see improved through the use of GenAI?
AI use-cases
Initiatives
Over 70% of enterprises report having >10 GenAI initiatives underway.
Q: How many individual GenAI initiatives are currently in proof-of-concept or production at your company?
The types of initiatives being implemented vary, with agents now a plurality.
Q: Roughly what percentage of your GenAI-related time or effort is allocated across the following initiatives (%)
Over half of respondents reported their fastest GenAI initiative taking longer than 8 weeks from idea to production.
Q: How many weeks did your fastest GenAI initiative take to go from initial idea to first production use?
Dreams
Ambitions span from fully automating customer-facing and compliance tasks to augmenting employee productivity in sales, development, and knowledge work.
Departments
Enterprises are dedicating a greater % of their budgets to workflow automation projects, beyond chatbots and knowledge management.
Software engineering and customer support are the two largest recipients of GenAI investment, but all departments have meaningful representation.

Internal projects (IT) investment is roughly 10% and they aim to improve efficiency. They focus on deployment of existing solutions. The rest of the budgets are focused on solutions for our customers via docs & KB chatbot and in product solutions.

T is driving bulk of the use cases in IT Ops and software development. The ROI is much.. easier to capture. Sales and marketing is another department that is spearheading a lot of use case to either upsell to customers or develop new ways to roll-out products.
Agent Themes
Over 70% of enterprises report having >10 GenAI initiatives underway.
ROI
On balance, the ROI is considered to be positive across the sample (+25pp). Most companies that do not yet see ROI are positive that it will realised in the near future.
Q: How would you categorise the return on investment (ROI) on your GenAI initiatives in production today?

Early measurements indicate that employee productivity is increased 30%.

ROI is still poor because most GenAI deployments are early-stage pilots that carry set-up costs – custom integrations, data cleansing, and added compliance reviews – without yet delivering large efficiency gains.

Strong ROI in apps development and testing, customer sales and service, underwriting productivity, fraud detection and claims processing.

“We are spending a great deal of time, resources, AI tech costs, changing processes and hiring the experts that ROI is yet to be realized. However, I strongly believe that we should be able to reap the benefits soon.
GenAI is most affecting costs today, with employee productivity and revenue growth secondarily.
Q: Which KPIs does your GenAI ROI most closely align to?
Concerns
Security is the dominant concern when deploying GenAI, far outweighing other factors. A shortage of skilled staff is the second most cited challenge. Despite large AI budgets (see ‘AI Procurement’), proving ROI and managing costs remain major anxieties.
Q: Please rank your top three business concerns about GenAI
Weighted Average (normalised to 10)
AI tech stack
Barriers
App security is also the largest technical barrier to implementing GenAI application. Additional critical blockers involve lack of talent and poor data quality. Regulation is the largest external blocker.
Q: On a scale of 0-10, what are the main barriers in productising GenAI applications in your organisation? (0 being not a barrier at all; 10 being barrier preventing progress)
Weighted Average (normalised to 10)
Capabilities
Many governance capabilities that they rate as urgent have not been put into production today.
Q: On a scale of 0-10, how important are the following capabilities for orchestrating GenAI usage at scale in your company? (0 being not at all important; 10 being absolutely critical). Which of these capabilities do you currently have in place today?
Models
>50% have >4 models.
OpenAI is the largest provider for more than half surveyed…
… representing >60% of total mean spend.
Q: Approximately what percentage of your overall GenAI usage or spend does this provider represent?
Tooling
>55% have >4 tools.
Enterprise GenAI budgets are fragmented: while OpenAI leads, significant spend still funnels to specialist platforms like UiPath for automation, Salesforce for CX, and Anthropic for safe language models.
Q: Please list out the top five tools by spend
Preference
They are open to a new, dedicated GenAI tool, and concerned about data quality. They need a platform that is enterprise-wide, secure, and cost-effective.
AI risk & compliance
Risks
Only a minority (23%) have cancelled a GenAI project.

Training data were incomplete, and remediation costs exceeded the projected benefits. Regulatory concerns over data residency and model transparency also surfaced.

A few proof of concepts did not result in a sufficient amount of net gain to warrant further investment.
Half would not even consider putting data into a public-cloud tool.
All have some form of shadow IT monitoring.
Enablers
Easy-to-use tools and sensitive data protections came up often as an important enabler for adoption.
# Times Mentioned in Top 3 (N=40)
AI policy
Ownership
>60% have a formal AI policy.

We have a very structured AI strategy driven by the top of the house. AI initiatives are governed centrally and managed through a committee.

Our AI implementation strategy is anchored on three core pillars: process automation, customer engagement, and productivity enhancement…

Encourage AI adoption in a well governed manner. AI use cases must be pre-approved and validated by senior leadership, compliance, security and governance leadership, observability and audit controls are mandatory.
A majority held the CTO or engineering leadership accountable.
Half of the respondents self-reported themselves as “intermediate” acumen – understanding the basics, with only 3% self-reporting as “advanced”.
Approvals
External AI was the most prominent example of needing a human-in-the-loop.
Internal tools were more flexible.
Frequency
75% planned to re-evaluated models at least every quarter.
Most exec teams discussed GenAI every quarter.
Q: How often do you discuss GenAI within the exec team?
Most organisations prefer to move more slowly over sacrificing control.
AI procurement
Budget
Over half of respondents had a dedicated GenAI budget. Over 4 in 5 respondents allocate more than 2% of their annual IT budget.
Almost half allocated 2-5% of the IT budget to GenAI projects.
Most didn’t have a different budgeting cycle.
Approvals
Most budgets are managed centrally.
Around half of enterprises can green-light GenAI projects exceeding $50k without procurement sign-off.
The majority of final approvals sit with the commercial side of the business.
Most don’t require any additional steps for GenAI projects.

We do formal risk assessments of the app from an info sec perspective and sometimes perform penetration tests.

CEO engagement required for GenAI.
Disclaimer
Dawn Capital LLP is a limited liability partnership incorporated in England & Wales (with registered number OC321845) and is authorised and regulated by the Financial Conduct Authority.
This report contains highly confidential and proprietary information about Dawn Capital LLP, its business and personnel. As a result, this report is provided on a confidential basis. By accepting this document, you agree that you will not use any of the information contained therein, including that regarding the business, operations, financial condition, or trade secrets of any of the investment partnerships managed or advised by Dawn Capital LLP (the “Funds”) and/or their respective portfolio companies, except with respect to your participation as an investor in said Funds.
This report is provided for information and discussion purposes only and does not purport to be comprehensive. No representation is made, or assurance given, that the statements, projections or forecasts are correct or that the objectives of the Funds and/or the portfolio companies will be achieved. No representation, warranty, express or implied, is given by any person as to the accuracy or completeness of the information and opinions contained herein and no responsibility or liability is accepted for the accuracy or sufficiency of any of the information or opinions, for any errors, omissions or misstatements or for any other communication, written or otherwise, in connection with the Funds and/or the portfolio companies. No information in this report will form the basis of any contract or an inducement to enter into any contract.
Accordingly, to the maximum extent permitted by law, neither Dawn Capital LLP nor the Funds and/or the portfolio companies nor any of their respective officers, employees, members or affiliates shall be liable for any direct, indirect or consequential loss or damage suffered by any person as a result of relying on any information or opinions contained herein or in any other communication in connection with the Funds and/or the portfolio companies. Accordingly, no reliance whatsoever should be placed on this report or its contents.
NO PERSON MAY DISCLOSE OR REPRODUCE ANY PORTION OF THIS REPORT WITHOUT THE PRIOR WRITTEN CONSENT OF DAWN CAPITAL LLP.