When Man Group announced its side-by-side partnership with Anthropic, the signal was clear: enterprise AI priorities have shifted. The collaboration focuses on building Claude Skills and agentic coding tools that move beyond chat-based assistance into AI executing workflows across investment operations. 

This marks a broader pivot across finance and asset management—away from AI responding to prompts and toward autonomous AI agents coordinating real work. 

Across trading desks, research teams, and risk functions, AI in investment workflows now means portfolio rebalancing, compliance checks, research synthesis, and execution pipelines running with limited human input. The question facing leadership teams is no longer “Can AI answer this?” but “Who controls how multiple agents act together?” 

Organizations preparing for this shift can explore structured enablement models through the AI CERTs Authorized Training Partner (ATP) Program 

AI Beyond Q&A: What Agentic Orchestration Actually Means 

Early copilots focused on productivity nudges—code suggestions, document drafts, query responses. Agentic AI operates differently. These systems plan tasks, call tools, validate outputs, and pass work to other agents. 

In finance, this already includes: 

• AI-driven investment execution that translates research signals into trades 

• AI-powered coding agents that write, test, and deploy internal analytics tools 

• Enterprise AI agents handling reporting, reconciliation, and compliance review 

The risk now surfacing across enterprises is Agent Sprawl—teams launching disconnected agents that duplicate work, create audit gaps, and raise accountability questions. 

Orchestration means defining: 

• Role boundaries between agents 

• Approval thresholds for autonomous actions 

• Shared memory and data access controls 

Without trained leaders, workflow automation with AI turns chaotic fast. 

Build orchestration capability early through certified AI training programs aligned with enterprise deployment needs. 

Agent Sprawl Is a Workforce Problem, Not a Tool Problem 

The anxiety around autonomous AI agents is real. According to the World Economic Forum, 44% of worker skills will change by 2027, with finance and technology roles seeing the highest transformation pressure. 

This fear often gets framed as displacement. Data shows a different pattern. 

IBM reports that organizations investing in formal AI training see 30–40% higher internal role mobility, reducing layoffs during automation cycles. 

Agentic systems require: 

• Oversight roles 

• Validation specialists 

• Workflow designers 

• AI governance leads 

These jobs emerge only when people are trained to manage AI executing workflows, not when AI is deployed in isolation. 

Can training partnerships mitigate job displacement concerns? 

Yes! When training connects directly to deployment. 

Standalone courses fail because they don’t map to real systems. Training partnerships succeed when tied to: 

• Live enterprise use cases 

• Role-based certification paths 

• Measurable job transitions 

A study by PwC found that employees offered structured AI upskilling are 3x more likely to stay during automation rollouts. 

This is where authorized training partner models matter. They link curriculum, certification, and employer demand. 

Organizations, institutes, and consultancies can become a partner through the AI CERTs ATP model to align skills with real agentic deployments 

How should institutions and companies collaborate to reskill workers at scale? 

The most effective reskilling systems run on three coordinated layers: 

1. Enterprises define execution roles 

Companies deploying AI for asset management must outline who supervises agents, who audits outputs, and who controls escalation. These roles become the target for training. 

2. Institutions deliver applied credentials 

Academic and professional bodies translate those roles into modular certification paths. The AI CERTs Authorized Academic Partner model enables universities to embed agentic AI coursework tied to industry needs. 

3. Governments support access and scale 

Public funding tied to verified certifications—not attendance—has shown higher employment outcomes. Singapore’s SkillsFuture model reports 70% placement or role change after AI-related credentials. 

This structure moves reskilling from intention to outcome. 

Industry associations can extend reach through the AI CERTs Association Partner pathway 

Why Finance Is Becoming the Blueprint for Agentic AI

The Man Group Anthropic partnership matters beyond asset management. Finance faces the strictest audit, risk, and compliance constraints—making it a proving ground for Generative AI in finance done responsibly. 

Claude Skills and agentic coding tools are being built with: 

• Traceable decision paths 

• Human override checkpoints 

• Version-controlled execution 

As Man Group CIO Rob Van Bruggen stated, “The value comes from AI acting within defined boundaries, not free-form experimentation.” 

Other sectors from healthcare to manufacturing—are watching closely. 

The Leadership Gap No One Is Talking About 

Agentic orchestration fails without trained decision-makers. The shortage is visible: 

• Gartner reports only 18% of enterprises have leaders prepared to manage multi-agent systems 

Training leaders to oversee enterprise AI agents is now a board-level issue. 

Consulting firms and learning providers can expand offerings through the AI CERTs Affiliate Partner model 

The Shift Is Already Underway 

AI answering questions was phase one. AI executing workflows defines phase two. The winners will not be those with the most agents but those with people trained to guide them. 

Agentic orchestration is a workforce strategy before it is a technology strategy. Organizations that invest early in structured partnerships, certifications, and shared responsibility models will convert anxiety into readiness and execution into advantage. 

Recent Blogs