🥶 Agentic AI is already changing the workforce - Stave, Kurt and Winsor

In Agentic AI Is Already Changing the Workforce Stave, Kurt and Winsor set out the major shift in outlook that the development of Agentic AI requires.

They suggests that organisations must actively shape how AI is integrated into its workforce strategy in order to:

• retain control over how AI is sourced, structured, and regulated in the enterprise, and to avoid unintended consequences.
• attract top candidates who increasingly expect AI support workflows.
• keep pace with other organisations embedding AI into their operating models.
• better comply with future requirements for auditable AI policies and governance frameworks (especially important for companies providing services to others)

The article outlines seven critical actions:

1. Map work tasks and outcomes.
   1. Deconstruct each role or project into its component tasks and outcomes to identify the tasks that can be better, faster, or more cost-effectively handled by AI agents and which are more effectively undertaken by a human or a hybrid approach.
2. Assess AI capability.
   1. To understand which AI models and platforms align best with specific tasks and workflows, create a capabilities catalogue: build an internal taxonomy of AI capabilities that map to roles. This avoids paying for systems that do not provide the best fit, or trying to leverage too much value from an unsuitable system.
3. Integrate your hybrid team.
   1. Define role boundaries: define which tasks AI will own, which tasks people will own, and how the escalation of problems should happen.
4. Redesign your business (and workforce) model.
   1. Consider multi-tiered models such as client-owned digital labor (you license or build your own AI solutions, effectively bringing “digital employees” in-house); leased digital labor (you “rent” AI agents from a third party, in a manner akin to traditional temp staffing); and fully outsourced AI subdepartments (you partner with a vendor that runs entire processes, such as order fulfillment or call centers, using both AI and a small team of human experts).
   2. Align these models with your financial, compliance, and strategic needs.
5. Set legal and ethical ground rules.
   1. Proactively address bias, liability, data governance, and broader societal implications of AI-led work. Work with legal, compliance, and ethics teams to draft enterprise-wide standards for AI usage to define whether and how AI learns from proprietary data, how to detect and remedy bias, and how to safeguard personal or sensitive information.
   2. Global enterprises may require different regulations in different areas.
   3. Don't overlook the importance of safeguarding ethical and legal rules and policies against unintended AI breaches.
6. Capture value continuously as it evolves.
   1. Constantly monitor performance, measure outcomes, and refine your AI-human mix. The required mix and approach will change as AI develops.
   2. Establish feedback loops that measure performance, update AI training data, and revise your sourcing strategies.
7. Remain human-centric.
   1. Whilst AI takes on most of the mundane, routine tasks, allow employees to carry out the high-value, human led tasks. This not only sustains morale but also delivers differentiating value to your enterprise. The human element is something that your competitors can’t simply download.
   2. Invest training and skill development to enable employees to not only adapt to working alongside AI but also leverage AI to amplify their own impact. Focus on capabilities like relationship building, ethical decision-making, and creativity—areas where humans still have a distinct edge.