Trust and human-AI collaboration set to define the next era of agentic AI, unlocking $450 billion opportunity by 2028

Core Insights - Agentic AI is projected to generate up to $450 billion in economic value by 2028, but only 2% of organizations have fully scaled deployment, with trust in AI agents declining [2][8][10] - Human oversight is deemed essential, with nearly 75% of executives believing its benefits outweigh costs, and 90% viewing human involvement in AI workflows positively [2][3][9] - Trust in fully autonomous AI agents has significantly decreased from 43% to 27% in the past year, with many executives concerned about the risks [5][8] Adoption and Implementation - Organizations are in early stages of agentic AI application, with 14% having begun implementation and nearly a quarter launching pilots [3][11] - 93% of business leaders believe scaling AI agents will provide a competitive edge, yet nearly half lack a strategy for implementation [3][10] - The report indicates that organizations with scaled implementation could generate approximately $382 million on average over the next three years, compared to around $76 million for others [10] Trust and Transparency - Trust in AI agents increases as organizations move from exploration to implementation, with 47% of those in the implementation phase reporting above-average trust [6][12] - Organizations are prioritizing transparency and ethical safeguards to enhance trust and drive adoption [6][9] Human-AI Collaboration - Over 60% of organizations expect to form human-agent teams within the next year, indicating a shift in perception of AI agents from tools to active team participants [7][9] - Effective human-AI collaboration is projected to increase human engagement in high-value tasks by 65%, creativity by 53%, and employee satisfaction by 49% [9][10] Challenges and Readiness - 80% of organizations lack mature AI infrastructure, and fewer than 20% report high levels of data readiness, indicating significant challenges in scaling agentic AI [12] - Ethical concerns, particularly around data privacy and algorithmic bias, remain prevalent, with only 34% of organizations actively addressing privacy issues [12]