Dem Hype Glauben schenken: Reale Daten zur Annahme und Wahrnehmung von generativer KI

Technology   |   Heather Ferguson   |   May 8, 2024 TIME TO READ: 11 MINS

There’s no doubt, generative AI has gone mainstream — and in less time than it takes a child to enter preschool. Since OpenAI released ChatGPT in November 2022, the hype engine has been firing on all cylinders about the possibilities of generative AI — consensus is that it is one of the most game-changing pieces of technology in a generation.

Though we investigated generative AI through two pulse surveys over the summer, enough time has passed to revisit the topic of generative AI and see how adoption trends and perceptions have changed in the last nine months.

This time, we talked to more people. Instead of surveying only business leaders about genAI, we also incorporated members of the general public into our sample to see what people outside of enterprises think about this technology. We surveyed 2,000 IT and data leaders in enterprises and 3,000 members of the general public across 11 countries. The goal was to discover where businesses are in their adoption and maturity around generative AI and how this compares to the public’s perception.

Let’s dig in.

Generative AI – More than just hype?

Much has been said about generative AI reaching the peak of the “hype cycle,” a Gartner research term describing the maturity of technology adoption.

The hype cycle is characterized by a peak at the top called the “peak of inflated expectations,” followed by the “trough of disillusionment,” where interest wanes in the technology and implementations fail to deliver.

In August 2023, Gartner published a report stating that generative AI was at the peak of inflated expectations. This means that at this point, we may expect to see indications that genAI is heading towards the trough.

Our research paints a different picture.

Seventy-eight percent (78%) of respondents feel that generative AI currently adds value to their organization — 43% say it adds “significant” value (Figure 1). This is higher than previous data, where only 34% of respondents said they saw substantial benefits from generative AI (Figure 2).

This bar graph titled "Current perception of the value of generative AI" depicts the opinions of business leaders on how valuable generative AI is. The graph is divided into four categories: "Generative AI adds significant value" (43%), "Generative AI adds some value" (36%), "Generative AI adds limited value" (18%), and "Generative AI adds no value" (4%). There is a cumulative total shown at the top of the first two categories, indicating that 78% of respondents believe generative AI adds significant or some value.

Figure 1 – What is your current perception of generative AI in terms of providing value for your organization? (Alteryx, May 2024)


This image displays a horizontal bar graph titled "Views on realizing the benefits from generative AI". The graph shows the distribution of responses from a survey regarding the benefits realized from using generative AI. The responses are categorized as follows: "We have realized substantial benefits of generative AI" (34%), "We have realized modest benefits of generative AI" (55%), "We have not realized any benefits of generative AI so far" (10%), and "It's too early to say" (1%).

Figure 2 – To what extent has your organization realized benefits from generative AI? (Alteryx, August 2023)

Likely because of this increased value generation, 62% of respondents plan to increase their investment in generative AI moving forward. In anonymous interviews with random respondents, one explained that their investment increase was driven by more departments becoming interested in the technology’s possibilities.

As more teams become aware of generative AI's capabilities and potential to streamline processes, increase productivity, and drive creativity, they naturally desire to use it.
Head of IT, Technology Company, USA

Pilot projects are successful, perceived risk is reducing

The increase in investment may also be tied to the fact that companies have found that generative AI has proven more straightforward to implement than they expected: 55% say it has been easier than expected to leverage generative AI in their organization; only 25% said it has been harder than they expected. (Figure 3) On average, businesses reported running three generative AI pilots since the start of 2023, and 77% of those pilots have been successful.

A bar graph titled "Ease of leveraging generative AI in organization" shows that 55% of respondents found it easier than expected, 25% found it harder than expected, and 20% found it to be what they expected. Source: Market Research, Attitudes, and Adoption of Generative AI, Alteryx, 2024.

Figure 3 – Has it been easier or harder than expected to leverage generative AI in your organization? (Alteryx, May 2024)

When it comes to the potential pitfalls of generative AI, it doesn’t seem like those are playing out as expected. Only 17% of respondents rate generative AI as high or very high risk to their organization, with most rating it as either low (31%) or moderate risk (34%).

Responses also paint a rosy picture regarding misuse of generative AI — 44% of businesses have not experienced any misuse of generative AI, and 48% have experienced misuse that has not significantly impacted the business. Only 5% reported experiencing generative AI misuse that had a significant negative impact.

What can explain these results? The skeptic in me speculates that companies have selected straightforward, low-risk use cases for piloting generative AI at this stage. As maturity increases and adoption reaches into more challenging use cases, results will start to reflect some of the challenges we typically see in the trough of disillusionment.

An optimist may speculate that generative AI won’t follow the traditional hype cycle path because of the nature of the technology — it’s easy to use, accessible, and natural language-friendly. Maybe the trough of disillusionment will be shallower or shorter with genAI.

Some responses support the optimist’s view:

The maturation of generative AI technology will lead to increased reliability and accessibility, making it more practical for financial institutions to adopt and integrate into their operations.
Chief Information Officer, Financial Services, UK.

The explanation likely lies somewhere in the middle: Perhaps some use cases are particularly well-suited to generative AI, and those will become standard practice very quickly, without impediment. At the same time, more complex applications may follow the traditional path of the hype cycle more closely, passing through the trough of disillusionment or falling off the path entirely.

Generative AI adoption and use – in business

Results paint a picture of rising genAI adoption, increasing maturity, and a growing impact on peoples’ roles across the business. On average, respondents estimate that 42% of knowledge workers in their organizations currently use generative AI in their roles.

Generative AI access

Regarding access, organizations are more likely to provide some form of open access to genAI than a controlled access version; 34% of organizations surveyed provide open access to genAI for all employees, and 40% provide open access for some employees. In comparison, 19% provide controlled access for all employees, and 8% provide controlled access for some employees. Related, 77% of respondents felt that employees currently have the right level of access to generative AI, and only 4% believe that employees need less access. (Figure 4)

This bar graph titled "Views on amount of access to generative AI tools" presents the opinions of respondents about the level of access employees have to generative AI tools within their organization. The graph shows three categories: "Yes, employees have the right level of access" with 77%, "No, employees require more access" with 19%, and "No, employees require less access" with 4%

Figure 4 – Do you feel employees have enough access to generative AI tools in your organization? (Alteryx, May 2024)

Generative AI usage: Data analysis leads the pack

For usage, respondents reported that the most common departments using generative AI were IT (55%), business intelligence/analytics/data science/data engineering (53%), customer support/service (46%), and marketing/communications (44%).

For use cases, the findings are similar to results from last summer, with one exception. The top reported use cases include:

  • Data analysis  – 43% (consistent with previous findings at 43%)
  • Cybersecurity – 37%
  • Customer support – 34%
  • Code generation – 32% (similar to what we saw previously at 31%)
  • Financial forecasting – 32%
  • Text generation – 32% (surprisingly lower than what we saw in August at 46%)

Interestingly, 66% of business leaders surveyed stated that their job responsibilities have already changed since the advent of generative AI, and 67% report that their team job responsibilities have also changed. (Figure 5)

This image shows two bar graphs side-by-side, titled "Changes in job responsibilities since the advent of generative AI." Each graph measures the extent to which job responsibilities have changed, for "Own job responsibilities" on the left and "Team’s responsibilities" on the right. For individual job responsibilities, the categories and their respective percentages are: "To a great extent" (37%), 4 (28%), 3 (17%), 2 (11%), and "To no extent" (6%). The cumulative total for the top two categories ("To a great extent" and 4) is 66%. For team responsibilities, the breakdown is: "To a great extent" (35%), 4 (32%), 3 (21%), 2 (9%), and "To no extent" (3%). The cumulative total for the top two categories here is 67%. Each graph uses a scale from "To a great extent" to "To no extent" to represent the impact of generative AI on job responsibilities, showing a majority reporting significant changes.

Figure 5 – To what extent have your and your team’s job responsibilities changed since the advent of generative AI in your organization? (Alteryx, May 2024)

Generative AI innovation

One of the more considerable changes we saw in the data over nine months is who leads generative AI innovation at companies. Forty-seven percent (47%) of respondents report that IT teams are leading generative AI innovation in the organization, 24% report the board leads innovation, and 21% rely on the C-Suite. (Figure 6)

This bar graph titled "Responsibility for generative AI innovation" illustrates the distribution of responsibility for generative AI initiatives across different leadership roles within organizations. The graph shows four categories with their respective percentages: "IT Leadership" (46%), "The board" (23%), "C-Suite" (21%), and "LOB heads" (Line of Business heads) (9%). Each bar represents the proportion of respondents who believe that the specified group holds responsibility for driving generative AI innovation within their organizations. The graph highlights that the majority of this responsibility currently lies with IT Leadership.

Figure 6 – Which individual or group are leading generative AI innovation within your organization? (Alteryx, May 2024)

This is significantly different from what we saw previously, when there was much less consensus on who owned generative AI, and the CEO was the highest influencer of genAI strategy, at 30%. (Figure 7)

This horizontal bar graph titled "Influence driving forward generative AI strategy" illustrates the distribution of influence among various leadership roles within organizations regarding the advancement of generative AI strategies. The graph presents percentages next to roles: "CEO, President, Owner" holds the most influence at 30%, followed by "VP, Director, Head of IT" at 25%, "CDO, CDAO, Chief Data Scientist" at 22%, and both "Head of AI, AI Analytics" and "CFO, Head of Finance" each at 19%. A smaller segment indicates that "There is no single person driving forward generative AI in our organization," which accounts for 2%. The graph effectively highlights the key decision-makers and their level of impact on generative AI initiatives.

Figure 7 – Who has the most influence driving forward generative AI strategy within your organization? (Alteryx, August 2023)

This is likely a good development, and it makes sense; IT leaders have distinct advantages in spearheading genAI innovation. They possess deep knowledge of the technology, its benefits and obstacles, the company’s current tech setup, and of the strategic direction that technology enables for the business.

It’s possible that in genAI’s infancy, the CEO was tasked with evaluating and starting the charge for generative AI implementation. Then, as companies’ genAI usage matured, CEOs have (rightfully) handed ownership of generative AI to their best-positioned leaders in IT.

Perceptions of generative AI – public vs. business

Some of the most interesting results from this research are how business users of generative AI differ in perspective from the general public. From awareness to emotions, business users and the general public paint different pictures of the technology.

Generative AI Tool Awareness
Business Leaders General Public
1.        ChatGPT (82%)

2.        Google Gemini (73%)

3.        Microsoft Copilot (70%)

4.        Grammarly (64%)

5.        DALL-E (56%)

1.        ChatGPT (69%)

2.        Microsoft Copilot (42%)

3.        Google Gemini (36%)

4.        Grammarly (32%)

5.        DALL-E (16%)

AI hallucinations come with the territory

There are significant differences in perception between business leaders and the general public regarding “AI hallucinations,” a response from a generative AI system that is false or misleading but presented as fact. Only 29% of the general public had heard of AI hallucinations, while 55% of business leaders said that they believed their employees were either extremely familiar (20%) or very familiar (25%) with the concept.

When it comes to experiencing AI hallucinations, 48% of business leaders state they have experienced misinformation produced by genAI, while only 37% of the general public have experienced it. AI hallucinations had a greater chance of negatively impacting the general public’s view of genAI, with 73% claiming that hallucinations had a detrimental effect on their trust in generative AI tools. On the other hand, business leaders were less inclined to believe that it had a negative impact on their trust, with only 65% of leaders expressing the same sentiment.

GenAI sentiment: Optimism with an undercurrent of regulation

Overall, business leaders were more emotionally optimistic about generative AI than the general public:

  • 89% of business respondents selected a positive emotion (interested, excited, curious, enthusiastic, or captivated), compared to 76% of the general public.
  • The general public was more likely to select negative emotions, with 61% of the general public selecting an emotion like skeptical, worried, apprehensive, fearful, or angry, compared to 44% of business leaders.

The general public is cautiously positive about generative AI and its transformative power (22%), but call out the need for it to be regulated (44%).

Regarding job displacement, a potential that often negatively impacts generative AI perception, 65% of business leaders felt that it was either highly likely (26%) or very likely (39%) that genAI would replace existing job roles in their organization. In comparison, only 35% of the general public agreed that generative AI would lead to job losses.

The quickly shifting landscape of generative AI

These results illustrate a shift in genAI adoption and maturity, showing that the technology has moved beyond just hype to becoming a valuable business tool. Businesses are not only adopting generative AI at an increasing rate but are also reporting greater benefits, with a majority planning to increase their investment in the technology. This is echoed by decreased perceived risks and a higher-than-expected rate of ease of use in successful pilot projects.

The diverging perceptions between business leaders and the general public highlight an interesting dynamic in which familiarity seems to breed (over?) confidence. While business leaders are more aware and appreciative of the capabilities and potential of generative AI, the general public, though positive, remains more cautious and concerned about the technology’s implications.

As genAI usage matures, more complex use cases may follow the traditional hype cycle, but early results are promising. Respondents’ blend of optimism and realism shows that while genAI may face challenges, they may be less than initially expected, and the potential for innovation and transformation will make up for any bumps in the road.