![\"\"](\"https:\/\/www.druckerforum.org\/blog\/wp-content\/uploads\/Sardas_Florence_1200x630px-1024x538.jpg\")
<\/figure>\n\n\n\nAs AI agents become part of the workforce, they\u2019re challenging established ideals about what good management and business value looks like. From defining outcomes and making judgement explicit to breaking down complexity, designing effective systems and creating stronger feedback loops, AI brings the fundamental disciplines of management back into focus to make us better managers of people.<\/p>\n\n\n\n
For decades, particularly in knowledge-based businesses, progression into management followed a familiar logic. People accumulate experience, develop technical expertise and become accomplished enough to guide and review the work of others. This is where AI challenges the equation of expert and manager.<\/p>\n\n\n\n
An AI agent can bring together information from multiple sources, process it at speed and apply it across various tasks. That doesn\u2019t make human expertise redundant. It simply changes where human ingenuity matters most. As businesses prepare for what\u2019s next, a manager\u2019s value (particularly the next generation) increasingly lies in applying judgement, providing context, challenging the technology\u2019s output and directing it towards the right outcome.<\/p>\n\n\n\n
Manage complexity, not tasks<\/strong><\/p>\n\n\n\n Ask an AI agent to \u201cbuild a transformation strategy\u201d and most of the important questions remain unanswered. What needs to change? Why? What are the constraints and acceptable trade-offs? What decision should the strategy ultimately enable?<\/p>\n\n\n\n In AI, this is called decomposition. In management, it\u2019s nothing new; capable managers turn that complexity into a sequence of work. The difference is AI makes the value of clarity more visible and, if clearer objectives, better context and more purposeful direction improve agents\u2019 performance, the same discipline should be applied to people using AI.<\/p>\n\n\n\n This leads to a deeper shift in how we work and, ironically, revives one of businesses oldest concepts: management by objectives. Where, traditionally, management easily become focused on activation, agentic AI now pushes managers towards outcomes.<\/p>\n\n\n\n Management thinker Peter Drucker argued that defining managers simply through responsibility for subordinates was too narrow, particularly in knowledge organisations. His thinking instead placed responsibility on the application and performance of knowledge. That distinction feels increasingly relevant. When access to knowledge becomes abundant, managerial value shifts. The manager\u2019s role doesn\u2019t require them to have all the answers, but they can deliver the best results by creating conditions for good answers to emerge. Judgement, rather than knowledge alone, has become the premium.<\/p>\n\n\n\n I see the same shift across the broader transformation landscape. As sophisticated AI becomes more accessible, simply having the technology will become less of a differentiator. Knowing where AI can genuinely improve the business, where it cannot and how capabilities can be translated into better decisions and outcomes will hold the greatest value for businesses and leaders worldwide.<\/p>\n\n\n\n Make institutional knowledge more valuable in the era of AI<\/strong><\/p>\n\n\n\n Some of the most valuable knowledge in an organisation has never appeared in a process manual. It sits with experienced people who know when to make an exception, escalate an issue, what good looks like in practice, or when the technically correct answer may not be the right fit for a client or customer.<\/p>\n\n\n\n AI brings this often-overlooked expertise into focus. For agents to work effectively, organisations need to understand the judgement behind their decisions, not simply document the steps in a process. In turn, this creates an opportunity beyond AI to capture how experienced people think, make valuable knowledge more accessible, help new colleagues learn and allow established ways of working to be challenged and improved.<\/p>\n\n\n\n From manager to architect: design human judgement alongside AI efficiencies<\/strong><\/p>\n\n\n\n As AI agents move from isolated tasks into wider workflows, managers need to look beyond individual outputs to the system around them. That means deciding where agents can act independently, where human judgement is essential, how work moves between the two and what happens when something goes wrong.<\/p>\n\n\n\n This is where governance and good management converge and essential to designing work well. A manager should become less of a supervisor and more of an architect, creating the conditions for good decisions and effective performance regardless of the system they\u2019re operating or process followed. <\/p>\n\n\n\n Convert good management into learning loops<\/strong><\/p>\n\n\n\n The most important lesson comes from what happens after the first output. As we know, successful AI systems improve through evaluation, feedback, correction and iteration. The same is true of successful teams where quality and optimisation are assessed. Feedback then informs the next attempt and performance improves because learning has been designed into how work is done.<\/p>\n\n\n\n This is management in one of its purest forms. As organisations bring humans and agents together, managers increasingly need to understand how each performs independently and how the combination works. There are three fundamental questions to optimise returns on human and technology investments:<\/p>\n\n\n\n This is also where the conversation about AI moves beyond efficiency towards genuine transformation. We are moving from building AI capacity to embedding AI capability into workflows where its value can be understood, measured and scaled. Questions about how we organise people, decisions and processes need to be embedded into what the technology can do, so that it does something valuable, but this is one of the biggest, long-term challenges.<\/p>\n\n\n\n Treat the next management challenge as an opportunity<\/strong><\/p>\n\n\n\n Learning to manage AI can make us better leaders and managers. It doesn\u2019t diminish the value of expertise; it changes where expertise counts. When information can be gathered at extraordinary speed, managers create value through judgement: framing the problem, setting direction, recognising what technology has missed, knowing when to intervene and building systems in which people and technology perform well together.<\/p>\n\n\n\n None of this is new and that\u2019s the point. For all the discussion about how radically AI could change management, one of its most useful effects may be to remind us what management was meant to do. We may learn to manage machines and, in the process, become better at leading people.<\/p>\n\n\n\n About the author:<\/strong><\/p>\n\n\n\n Florence Sardas, Chief Transformation Officer of Forvis Mazars Group<\/em><\/strong><\/em><\/p>\n\n\n\n She leads the firm’s global transformation agenda, spanning technology, innovation, AI and digital services. With extensive experience at the intersection of business transformation, professional services and emerging technologies, she works with leaders across the world to scale innovation, redesign operating models and accelerate the adoption of AI. Her focus is on how organisations can harness technology not only to improve performance, but also to strengthen expertise, trust and human collaboration in an era of profound change.<\/em><\/p>\n","protected":false},"excerpt":{"rendered":" As AI agents become part of the workforce, they\u2019re challenging established ideals about what good management and business value looks like. From defining outcomes and making judgement explicit to breaking down complexity, designing effective systems and creating stronger feedback loops, AI brings the fundamental disciplines of management back into focus to make us better managers of people.[\u2026]<\/a><\/p>\n","protected":false},"author":3,"featured_media":6150,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_seopress_robots_primary_cat":"none","_seopress_titles_title":"","_seopress_titles_desc":"","_seopress_robots_index":""},"categories":[396],"tags":[397,411],"_links":{"self":[{"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/posts\/6146"}],"collection":[{"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/comments?post=6146"}],"version-history":[{"count":1,"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/posts\/6146\/revisions"}],"predecessor-version":[{"id":6151,"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/posts\/6146\/revisions\/6151"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/media\/6150"}],"wp:attachment":[{"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/media?parent=6146"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/categories?post=6146"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.druckerforum.org\/blog\/wp-json\/wp\/v2\/tags?post=6146"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}\n