A Salesforce survey of nearly 6,000 global knowledge workers suggests that AI has a data problem. Nearly 6 in 10 AI users say it’s difficult to get what they want out of AI right now, with over half claiming they don’t trust the data used to train today’s AI systems. In fact, more than three-quarters of workers say that accurate, complete, and secure data is critical to building trust in AI.
Why it matters: Eight in ten business leaders believe generative AI will lower business costs and increase revenue. However, the majority of workers believe that trusted customer data is required for generative AI to be used successfully in their role and at their organisation. This new survey indicates that generative AI outputs must be grounded in trusted data to close this gap — and reap the benefits of the technology across the enterprise.
Salesforce perspective: “The future of enterprise AI isn’t about more data – it’s about the right data. When AI is grounded in a company’s own data, it delivers more useful results and ultimately drives greater trust and adoption.”– Wendy Batchelder, SVP, Chief Data Officer
The Salesforce research found:
AI lacks the data needed to be useful right now, putting adoption at risk
- 56% of AI users say it’s difficult to get what they want out of AI.
- 51% of workers say generative AI lacks the information needed to be useful.
- 75% of those who don’t trust the data that trains AI also believe that AI lacks the information needed to be useful.
- 68% of those who don’t trust the data that trains AI, are hesitant to adopt it.
AI trained on unreliable data creates a serious trust gap
- 62% of workers say out-of-date public data would break their trust in AI.
- 71% of workers say consistently inaccurate outputs would break their trust in AI.
- 54% of AI users don’t trust the data used to train AI systems.
- 68% of workers who don’t trust AI say the training data is unreliable.
With good data, AI can earn workers’ trust
- Workers cite the following as critical to building trust in AI:
- Accurate data (82%)
- Secure data (82%)
- Holistic/complete data (78%)
- 53% of workers say training AI on comprehensive customer/company data builds their trust in the tool.
