Why Human Strategy Matters as AI Agents Become More Autonomous

As artificial intelligence (AI) agents take on more tasks and decisions, companies need to think carefully about how those systems work together, argues TechRadar’s Marie Aiello.

AI tools can already analyze audiences, create and adapt marketing content, check it against requirements and evaluate results. Connecting these steps could allow one action to trigger another automatically. But a workflow can follow its instructions and still produce poor results if its goals, context or links between decisions are poorly designed.

For example, a system might target people likely to respond to an offer while overlooking customers the business wants to develop over time. A content agent could optimize for clicks in ways that do not fit the brand’s direction, or a performance system could recommend changes based on a metric that misses a broader business objective. Aiello argues these outcomes can stem from how the system was set up, rather than from the AI failing to follow instructions.

Companies have faced a related challenge with specialized business software. Teams often had to connect separate platforms, reconcile conflicting information and carry context between systems. Wider adoption of AI agents could reproduce that coordination problem, with agents using different data, instructions and goals.

Aiello says companies should decide where agents can act independently, where monitoring is enough and when a person should be involved. The appropriate level may vary by workflow and by the consequences of a decision. As AI models change quickly and become more widely available, she argues, an organization’s understanding of its operations, customers and goals may be harder to replicate than access to a particular model.