July 26, 2025
5 min read
nojitter.com
Forrester finds AI agents ready for enterprise use, but adoption hindered by mistrust, data gaps, and workforce readiness.
Forrester: AI Agents Are Ready, People and Data Are Not
The key technological components for truly autonomous AI are being assembled, but user mistrust in AI and missing usable data hamper adoption. In a report published on July 8, 2025, Forrester analysts detailed how the critical technology components needed for agentic AI applications — which can act autonomously on behalf of enterprises or individuals, perform tasks, make decisions, and interact with data or other systems — are coming together. However, the report also highlights significant non-technical barriers to adoption.Key Components for AI Agent Adoption
- Tool discovery and integration: Approaches like the Model Context Protocol (MCP) help AI agents discover and integrate tools.
- Agent-to-agent interoperability: The Agent2Agent protocol (A2A) aims to enable communication between AI agents.
- Orchestration capabilities: Systems that direct AI agents on what to do and provide user interfaces for human interaction.
- Low trust in AI outputs: Both employees and consumers remain wary of AI decisions and results.
- Misaligned workflows and missing data: Many organizations lack the clean, accessible data needed for AI agents to function effectively.
- Unclear and fragmented regulatory guidance: Regulatory uncertainty complicates deployment.
- Workforce readiness: Training employees to use AI tools effectively and overcoming fear of displacement is critical. Stephanie Liu, Forrester senior analyst and report co-author, emphasizes, “You have to ensure you're bringing employees on the journey. It's not just training on how to use the tool, but helping them get over the fear, uncertainty and doubt of learning to use that which may outsource or displace them.”
- Clearly define what they want their AI agents to do.
- Identify the data sources the AI needs to access.
- Start small by automating one step in a workflow and expand iteratively. Early experimentation helps build a roadmap for future AI agent capabilities.
- Intuit Adds AI Agents, Other Enhancements to Intuit Enterprise Suite
- AI-Driven Crypto Trading Tools Reshape Market Strategies in 2025
- How to Use ChatGPT Agent for Crypto Trading in 2025
- Exclusive: Sysdig and BitMEX Sound Alarm on AI-Powered Crypto Threats
Barriers to Broader Uptake
Forrester identifies several adoption challenges:Insights on AI Agent Evolution
Thanks to advances in large language models, AI agents can now "reason" and determine the next best step in workflows, using third-party tools and data autonomously if granted access. This contrasts with humans and traditional automation technologies like robotic process automation (RPA), which require manual programming and documentation of each step. Liu notes that formal documentation often does not reflect how tasks are actually performed. AI agents will eventually learn the most efficient ways to complete processes without requiring perfect workflows to be documented. The report highlights various AI agent use cases — including consumer engagement, employee support, and enterprise automation — each progressing at different rates. The technology is rapidly evolving from assistant- or copilot-style applications (e.g., summarization, writing assistance) to "solver agents" that execute tasks autonomously on behalf of humans or organizations.Recommendations for Organizations
Liu advises organizations to:Related reading: No Jitter Roll: Avaya Plans to Adopt Model Context Protocol
About the Author
Matt Vartabedian is Senior Editor at No Jitter, covering AI (predictive, generative, and agentic) as it relates to enterprise communications, including unified communications, contact centers, and digital workplaces. With a journalism career starting in the late 1990s and two decades as a cellular industry analyst, Matt brings deep expertise grounded in research and data analysis.Source: Originally published at No Jitter on July 25, 2025.