> ## Content Index > Fetch the complete content index at: https://www.nevillehobson.io/llms.txt > Use this file to discover other available public pages before exploring further. # When AI enters the org chart, who really holds the power? - URL: https://www.nevillehobson.io/when-ai-enters-the-org-chart-who-really-holds-the-power/ - Published: 2026-11-02T07:22:00.000Z - Updated: 2026-11-02T07:21:59.000Z - Description: AI promises to make organisations more productive, responsive and intelligent. But as it moves from assisting employees to influencing management decisions, what happens to communication, trust and human relationships? - Author: Neville Hobson - Tags: Artificial Intelligence, Communication, Strategy, Workplace, Future of Work It's been another extraordinary week for artificial intelligence. Anthropic's planned stock market flotation attracted enormous attention, with expectations of a valuation measured in trillions of dollars. Alongside the excitement about AI's commercial potential come warnings about its risks, including from some of the very people developing it. The contrast is striking. We're being invited to imagine a future of extraordinary abundance and productivity while simultaneously being warned that the technology making it possible could have consequences we don't fully understand. Amid all this, I came across a thoughtful essay published by OpenAI that asks a more immediate question. Not whether AI will transform society, but what happens when it becomes part of the organisations where we work. [*The future of the firm*](https://openai.com/index/the-future-of-the-firm/?ref=nevillehobson.io), by OpenAI's [Hemanth Asirvatham](https://www.linkedin.com/in/hemanthasirvatham/?ref=nevillehobson.io) and [Elliott Mokski](https://www.linkedin.com/in/elliott-mokski/?ref=nevillehobson.io), considers how AI might enter the organisational hierarchy. The authors are careful to describe their ideas as possible scenarios, not predictions, and note that they're expressing personal views rather than company policy. Their central proposition is intriguing: AI won't simply change the work people do. It could fundamentally change how organisations are structured, how authority is exercised and how decisions are made. For anyone concerned with organisational communication, leadership and trust, that deserves attention. ## **Three possible futures for the workplace** The authors suggest three ways AI might find its place in the organisational chart. The first is **beneath us**. AI agents take on routine work, effectively becoming assistants or teams reporting to individual employees. Everyone becomes something of a manager, directing AI resources to accomplish tasks that previously required substantial time or additional colleagues. It's an appealing prospect. Employees spend less time on repetitive activities and more on creative thinking, judgement and decisions. Even someone starting their career might command capabilities previously available only to senior colleagues. The second possibility is **beside us**. Here, AI operates alongside human employees as a colleague, perhaps taking responsibility for complete business functions. A manager might assign an entire marketing campaign to AI agents while human teams concentrate on other priorities. The third, and perhaps most provocative, is **between us**. AI becomes part of the management structure itself, analysing organisational activity, identifying problems and advising senior leaders. Some employees might effectively find themselves working under AI systems that coordinate their activities or assess their contributions. The authors believe elements of all three could emerge together. Indeed, they argue that early versions are already appearing. None of these scenarios necessarily means wholesale job losses, although the authors acknowledge that employment could decline as automation expands. Their interest is in how organisations might evolve as human and artificial intelligence become increasingly intertwined. It's an imaginative and often optimistic account of what might be possible. But I found myself wondering about something largely missing from these scenarios: the human relationships that make organisations work. ## **The promise of a more capable organisation** There's much to welcome in the possibilities the authors describe. Imagine employees freed from repetitive administrative work, able to focus on activities that require imagination, empathy and judgement. Imagine junior colleagues having the resources to develop ideas without waiting for budgets or additional headcount. Or consider senior leaders with a much clearer understanding of what's happening across their organisations. An AI system might identify a developing problem before it becomes a crisis. It could reveal gaps between departments, recognise contributions that managers have overlooked, or help employees make better-informed decisions. Some of the bureaucracy that slows organisations down might disappear. Teams could experiment more readily, learn from mistakes and adapt without navigating endless approval processes. The authors also raise an interesting possibility about employee autonomy. If AI can assess work outcomes more effectively, perhaps organisations could become less concerned about how, when, and where people work. Instead of measuring activity through crude indicators such as hours online or messages sent, managers might better understand actual contributions. That sounds attractive, particularly after years of arguments about remote and hybrid working. But greater productivity doesn't necessarily mean a better working life. If AI enables employees to complete tasks in half the time, what happens to the other half? Is it an opportunity for learning, creativity and more meaningful work, or simply an invitation to double their output? That's a leadership choice, not a technological inevitability. Another significant assumption behind this vision of greater employee autonomy is that increased organisational visibility will produce greater trust. I'm not convinced that's necessarily how things would work. ## **When visibility becomes surveillance** One of the essay's most striking ideas is that AI could act as a continuously available internal management consultancy. Imagine a chief executive able to ask an AI system what's happening across hundreds of projects, conversations and documents. The system identifies problems, highlights achievements and recommends action, perhaps before anyone has formally reported anything. That capability has obvious value. There's also an uncomfortable question: what happens when an organisation can observe, analyse and interpret almost everything its employees do? The authors acknowledge the possibility of surveillance but suggest that intelligent monitoring might actually give employees greater freedom. If managers can see that people are delivering good results, they may feel less need to control how those results are achieved. Perhaps. But monitoring doesn't become less intrusive simply because the technology behind it is more sophisticated. Employees who know their messages, conversations and contributions are continuously analysed may become more cautious about what they say. They might avoid challenging established thinking or expressing uncertainty. Informal exchanges where people test ideas, admit mistakes, and build trust could become more guarded. 💡 And who decides what constitutes a valuable contribution? Who determines whether an employee's judgement was sound, or whether a disagreement was constructive rather than disruptive? An AI system will reflect the objectives, assumptions and priorities embedded in its design and deployment. Those choices are made by people. Technology might give leaders unprecedented visibility into their organisations. Whether they use that visibility to empower employees or exercise tighter control remains a leadership decision. ## **The communication paradox** This is where I think the essay raises a particularly important issue for organisational communicators. AI could make information flow within organisations dramatically more efficient. Leaders might receive accurate summaries of complex discussions. Employees could find the information they need almost instantly. Teams might coordinate activities without endless meetings, reports or layers of management. In one sense, communication could become faster, more comprehensive and less expensive than ever before. 💡 Yet here's the paradox. Moving information efficiently isn't the same as communicating effectively. Communication isn't simply about transmitting information from one person to another. It's about creating shared understanding, negotiating meaning, building relationships and establishing trust. An executive might receive an accurate AI-generated summary of employee concerns without understanding why those concerns matter so deeply to the people expressing them. An AI system might identify disagreement between colleagues without appreciating that the disagreement reflects healthy professional debate. A leadership team might become exceptionally efficient at communicating decisions while becoming much less effective at involving people in making them. And an organisation could have more information about its employees than ever before while understanding them less well. That's a possibility worth taking seriously. The danger isn't necessarily that AI communicates badly. Increasingly, it may communicate remarkably well in the conventional sense of producing clear, persuasive and appropriately framed messages. The danger is that we confuse the quality of those messages with the quality of the relationships behind them. Trust isn't something an organisation can generate simply by improving its information systems. It develops through behaviour, experience, consistency and the belief that people are being treated fairly. Those qualities are considerably harder to automate. ## **Who makes the decisions, and who is accountable?** There's another challenge: if AI increasingly influences management decisions, where does accountability sit? Consider an AI system recommending that a project be cancelled, a team reorganised, or an employee's performance questioned. A human manager might retain formal responsibility for the eventual decision. But how meaningful is that responsibility if the manager cannot adequately assess the reasoning behind the recommendation? And what opportunities would employees have to question a decision influenced by an AI system whose workings they cannot inspect? The issue becomes particularly important when AI recommendations concern people rather than processes. Organisations already struggle with transparency and fairness in decision-making. Introducing systems that can influence decisions at scale could make those difficulties far more serious. There's also a question of power. The authors recognise that AI could strengthen centralised control by giving executives much greater insight into activities throughout their organisations. That's a significant possibility. An organisation might remove layers of middle management while concentrating authority more firmly at the top. Employees could gain sophisticated AI assistants yet have less influence over the decisions that shape their working lives. Greater technological capability doesn't automatically mean greater human agency. ## **A challenge for organisational communicators** What does all this mean for communication professionals? I think it reinforces an argument I've made before: communicators need to be involved in decisions about AI adoption, not simply brought in afterwards to explain what's changing. There's another dimension, too, highlighted by [James Allsopp](https://www.linkedin.com/in/jamesallsopp88/?ref=nevillehobson.io) in his article [*The Rise of the AI Employee: Are We Ready for Digital Colleagues?*](https://agile.org.uk/the-rise-of-the-ai-employee/?ref=nevillehobson.io) Employees shouldn't simply be told how AI will change their work. They should have opportunities to influence those changes. The people closest to the work often understand its complexities better than those making technology decisions. Listening to them isn't just good communication practice. It could make the difference between AI that genuinely improves working life and AI that introduces problems nobody anticipated. 💡 If AI begins influencing how work is allocated, how performance is assessed and how management decisions are made, these aren't merely operational or technological questions. They're questions about organisational culture, trust, fairness and legitimacy. Communicators should be asking difficult questions. How will employees know when AI is influencing decisions that affect them? What opportunities will they have to challenge those decisions? How will leaders explain the boundaries of AI authority? And how will organisations ensure that increasing technological efficiency doesn't diminish meaningful human participation? Perhaps most importantly, how will leaders maintain genuine dialogue with employees when machines can filter, summarise, and interpret so much organisational information? These questions belong alongside those being considered by technology teams, HR professionals, legal advisers and senior executives. Communication professionals have experience of how people respond to uncertainty, how trust is built or damaged, and why technically correct messages don't always produce understanding or acceptance. That expertise could become increasingly valuable, but only if communicators recognise that their contribution goes beyond making AI-driven change understandable. They should also help shape the conditions under which that change takes place. ## **The choice is still ours** I enjoyed reading *The future of the firm*. It's ambitious, provocative and refreshingly willing to consider several possible outcomes rather than present one inevitable future. Its three scenarios – AI beneath us, beside us, and between us – offer a useful way of thinking about how AI might reshape organisations. But they also raise questions that go well beyond organisational charts and productivity. 💡 What happens to human judgement when machines increasingly influence decisions? What becomes of trust when leaders can monitor almost everything? And how do organisations preserve meaningful relationships when information flows more efficiently than ever? Wherever AI sits, the technology won't determine whether an organisation becomes more humane, more trustworthy or more inclusive. Those outcomes will depend on the people making decisions about its use. The most important question may not be where AI belongs on the organisational chart, but what kind of organisation we want that chart to represent. --- ## **Sources:** - [OpenAI: The future of the firm](https://openai.com/index/the-future-of-the-firm/?ref=nevillehobson.io) – Hemanth Asirvatham and Elliott Mokski, 2026. - [Anthropic: Preliminary IPO registration statement](https://www.sec.gov/Archives/edgar/data/2130357/000162828026067168/anthropicpbc-sx1.htm?ref=nevillehobson.io) – US Securities and Exchange Commission, 26 October 2026. - [Reuters: Anthropic IPO stands alone as AI market jitters sideline others](https://www.investing.com/news/economy-news/analysisanthropic-ipo-stands-alone-as-ai-market-jitters-sideline-others-4966313?ref=nevillehobson.io) – 22 October 2026. - [Agile.org.uk: The Rise of the AI Employee: Are We Ready for Digital Colleagues?](https://agile.org.uk/the-rise-of-the-ai-employee/?ref=nevillehobson.io) – James Allsopp, 28 October 2026. (AI-generated image at top created with ChatGPT.)