80% of AI decision-makers express concern over data privacy and security.

80% of AI decision-makers express concern over data privacy and security.

80% of AI decision-makers express concern over data privacy and security.

Organizations are excited about the potential of generative AI to boost business and productivity. However, challenges like lack of strategic planning and talent shortages hinder them from fully leveraging its benefits, according to a recent study by Coleman Parkes Research, sponsored by SAS. The study, conducted in early 2024, surveyed 300 U.S. decision-makers in GenAI strategy or data analytics.

Marinela Profi, a strategic AI advisor at SAS, emphasized that large language models (LLMs) alone are insufficient for solving business challenges. Instead, GenAI should enhance hyper-automation and accelerate existing systems and processes, rather than being seen merely as a novelty. Profi advises that organizations develop a thoughtful strategy and invest in technologies that ensure the integration, governance, and explainability of LLMs before fully committing.

The study highlighted four major hurdles organizations face with GenAI implementation:

1. **Trust and Compliance**: Only 10% of organizations have effective systems to measure bias and privacy risks in LLMs. Furthermore, 93% of U.S. businesses lack a comprehensive governance framework for GenAI, putting them at risk of noncompliance with regulations.

2. **Integration Issues**: Many organizations struggle with compatibility issues when attempting to integrate GenAI with existing systems.

3. **Talent Shortages**: There is a significant gap in in-house GenAI expertise. HR departments face difficulties in recruiting suitable talent, leading to concerns among leaders about their ability to maximize their GenAI investments.

4. **Cost Predictions**: Leaders are challenged by the direct and indirect costs associated with LLMs. While initial cost estimates for model usage are provided, expenses related to private data preparation, training, and ModelOps management are extensive and complex.

Profi further noted that success with GenAI depends on identifying real-world use cases that offer the highest value and meet human needs sustainably and scalably. She reiterated SAS’s commitment to assisting organizations in investing wisely and staying resilient amid rapidly evolving AI technologies.

These findings were revealed at the SAS Innovate conference in Las Vegas, an event focused on AI and analytics for business leaders, technical users, and SAS partners.