In a survey of 300 decision makers on Generative AI adoption plans, many respondents underestimated the pre-requisites of successful implementation

In November and December 2023, 300 C-suite executives (70%) and vice presidents or directors (30%) from the Americas (17%), Europe (17%), and the Asia Pacific region (66%) were polled about how their organizations were implementing — or planning to implement — generative AI (GenAI) technologies and what barriers to effective deployment they faced.

Some 70% of respondents managed information technology, data, and data engineering-related functions. All respondents were distributed among industries including financial services, banking, and insurance (17%); consumer packaged goods and retail (17%); manufacturing and automotive (17%); technology and telecom (17%); logistics (13%); energy, oil, and gas (10%); and media and communications (10%).


Additionally, eight in-depth interviews were conducted with data and AI experts in the same period.

The survey data showed the following key findings:

    1. 60% of respondents expected GenAI to disrupt industries

    These are the ways in which a threat actor gains access to a system or resource. The most common approach is using a spear phishing technique via email. This typically includes either a malware attachment or embedded link to an external malware service that a user inadvertently clicks on. People are the weakest link in the security chain and constitute a preferred route to gaining access to systems and networks.

    2. 78% did not see AI disruption as a risk

    Rather than being concerned, these respondents saw GenAI as a competitive opportunity. Some 8% regarded it as a threat, while 65% cited that their businesses were actively considering new and innovative ways to use GenAI to unlock hidden opportunities from data.

    3. 9% companies went beyond experimentation with GenAI

    76% of respondents had worked with GenAI in some way in 2023, and 9% had adopted the technology widely. Among the rest who had experimented, deployment was in one or a few limited areas. The most common use case was automating non-essential tasks. Those that were deploying GenAI cited plans to “more than double” deployment this year. They expected to “frequently apply” the technology in customer experience, strategic analysis, and product innovation areas by end-2024. Meanwhile, respondents cited plans to increase use of GenAI in specific fields relevant to their individual industries, include coding for IT firms, supply change management in logistics, and compliance in financial services.

    4. A need to address IT deficiencies for GenAI

    Under 30% of respondents ranked IT attributes at their companies as conducive to rapid adoption of GenAI. Those with the most experience of rolling out GenAI had even less confidence in their IT. Of this group, 65% cited that their available hardware was, at best, modestly conducive to rapid adoption.

    5. Five factors impacting successful use of GenAI

    Respondents, both in general and AI early adopters, cited five non-IT impediments to the extensive use of generative AI.

      I. Risk: 77% of respondents cited their regulatory, compliance, and data privacy environment as a leading barrier to rapid AI adoption.

      II. Budgets:56% listed IT investment budgets as a leading barrier.

      III. Competitive environment: Early adopters of GenAI in the survey were more than twice as likely to see the competitive environment as an enabler of rapid adoption rather than as a barrier.

      IV. Culture:Early adopters of GenAI in the survey were more likely to regard openness to innovation as an enabler of rapid adoption.

      V. Skills:The skills needed for significant AI projects were in short supply; respondents that were early adopters were more acutely aware of the shortage of available talent.


The survey was sponsored by Telstra International, whose Director of South Asia and Head of Global Enterprise, Geraldine Kor, said: “Effectively deploying GenAI solutions is predicated upon having 100% confidence in the end-to-end operationalization of capturing, processing, contextualizing, and actioning data.”