Machine learning and predictive AI have been part of many businesses for years. However, the world’s reaction to generative AI, as introduced to them by ChatGPT, has been qualitatively different. ChatGPT was certainly not the first generative AI or LLM, but for most people, it was the first Large Language Model (LLM) available to try firsthand. ChatGPT put AI into a website that lowered the barrier to entry1 and suddenly, anyone could use one of the world’s most powerful LLMs. In the 14 months since its launch, ChatGPT has become synonymous with AI and has kicked off the corporate adoption clock.

Since ChatGPT’s launch in November 2022, we’ve seen historically fast AI adoption in the enterprise space. The hype around AI is astounding: 96% of workers believe generative AI can help them in their jobs2 and ChatGPT reached 100 million users two months after its launch3, setting the record for the fastest growing consumer user base.

On the other hand, AI is also frequently compared to Blockchain, which is still struggling to get traction within enterprises. We’ve heard cautionary tales about AI citing fake court cases, leaking sensitive data, and quickly introducing bias and discrimination to its outputs.

ivision surveyed 90 executives and got 60 responses (a 67% response rate!) and we wanted to share some anonymized data and insights we gained from the data.

As a baseline, we found that 72% of survey respondents are in the process of rolling out new AI solutions. An additional 17% are enabling AI in existing SaaS platforms, and 12% are doing neither. Budgets reflect roughly the same split: 70% of companies surveyed had at least some budget for AI pilot or project work in 2024, and 15% had budgets over $750k.

This pace of adoption is nevertheless incredible given the significant (and real) concern around errors and data privacy surrounding AI. Early cautionary tales reinforced core fears and few AI providers have been able to provide reassuring, trustworthy answers. Trust and privacy are by far the most important criteria to our customers: 42% of the respondents said that an AI backend run by a trusted partner matters most, and 40% said privacy matters most. That’s a total of 82% that selected either trusted partner or privacy as their top priority.

When asked their primary concern while integrating AI into their business processes, 42% said data privacy/security and 45% said hallucinations/errors. That comprised 87% of the total answers and paralleled two of the most prominent cautionary news tales: Samsung banning generative AI tools in May 2023 after employees leaked source code to ChatGPT and a judge sanctioning an attorney in June 2023 after his brief cited fake judicial opinions and legal citations.

Our survey also showed that most companies were willing to take calculated risks to roll out AI and reap the benefits. They are in the process of adopting the technology while they work on education and policy in parallel. Of the companies surveyed, only 10% said they had mature guidelines or policies to mitigate unauthorized exposure of data through AI. This has been a particularly challenging area for every company we’ve talked to, both anecdotally and reflected in this data.

However, these companies have also seen the tremendous value of AI. 83% of companies believe there is value or have been able to show that value. Of the remaining 17%, only 2% do not believe there is value and 15% have not made any investments to date. Similarly, 80% believe the ROI on AI will justify its cost. But how soon? 33% believe that time is now, 15% believe ROI will justify the cost in 6 months, and 32% predict results in 12 months.

Microsoft made a big bet early on as OpenAI’s largest investor and commercial partner since 2016 and it’s paying off. When Satya Nadella became Microsoft’s third CEO in 2014, Steve Ballmer offered him advice (and a challenge): “Be bold and be right”. Microsoft’s embrace of AI and methodical rollout certainly embody that challenge. Microsoft is far and away the most used and trusted partner among the respondents, with 53% of all respondents using Azure/Copilot. An additional 15% are using OpenAI. Only 3% each are using AWS/Bedrock and Google Cloud/Bard.

Overall, the survey shows that a large majority of companies and employees see value in generative AI and are rolling out AI at a historic pace. Companies are relying on their trusted partners to address core concerns around the technology as it matures, and they can show the ROI on these early investments. Microsoft and OpenAI are the primary beneficiaries of these rollouts.

Disagree? Have an AI use case to discuss with us? Ready to get started rolling out AI safely within your enterprise? Give us a call. We look forward to your feedback.

  1. Microsoft has now made ChatGPT available for use by anyone via Copilot. It’s free to use and, depending on your account type, can be linked to your WORK data or WEB via a toggle at the top of the page. ↩

  2. https://www.oliverwymanforum.com/global-consumer-sentiment/how-will-ai-affect-global-economics.html ↩

  3. https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/ ↩