If AI can replace engineers, isn't management even more automatable?

Most AI discussions seem to assume that software engineers are the first white collar workers to be replaced. I'm starting to think management may actually be more exposed. Here's why. Engineering is not just writing code. It's debugging messy production systems, handling undocumented behavior, working around hardware and infrastructure constraints, integrating imperfect APIs, and constantly adapting to edge cases. AI is getting very good at coding, but reliable execution in complex real world environments is still difficult. Management, by contrast, is largely an information processing and decision making function. A manager typically: * Gathers information from multiple teams. * Prioritizes work. * Allocates resources. * Assesses risk. * Tracks execution. * Resolves conflicts. * Communicates decisions. * Forecasts outcomes. * Sets strategy. These are all tasks that depend on processing large amounts of information, an area where AI is improving rapidly. An AI manager could theoretically: * Read every Slack message, document, code review, incident report, customer complaint, sales call, financial metric, and support ticket simultaneously. * Monitor thousands of KPIs continuously instead of relying on weekly updates. * Detect emerging risks earlier than humans. * Evaluate hundreds of strategic options before making a recommendation. * Apply consistent decision criteria instead of being influenced by office politics, hierarchy, fatigue, or recency bias. * Provide evidence for every recommendation. * Operate 24/7 across every time zone. * Instantly incorporate new research, regulations, market data, and technical knowledge. * Communicate with every employee in their preferred language and level of technical depth. Executives often talk about having the "big picture." A sufficiently capable AI could arguably have a much larger picture than any individual CEO because it can reason across the entire organization at once instead of relying on summaries passed through multiple management layers. If the argument is that engineers are vulnerable because coding is a cognitive task, then management seems at least as vulnerable, since it is almost entirely a cognitive and information processing role. The real barriers don't seem to be technical capability. They seem to be accountability, governance, incentives, legal responsibility, and whether organizations are willing to delegate high impact decisions to AI. What am I missing? Is management fundamentally harder to automate than engineering, or is the conversation focused on engineers simply because AI became useful for coding before it became useful for executive decision making?