Lenin Gali is a former CIO and currently the Chief Digital Officer at Atomicwork.
On June 1 this year, Deloitte eliminated traditional job titles for roughly 181,500 U.S. employees, retiring the analyst-consultant-manager ladder in favor of role-specific titles organized by job family.
Accenture ran an $865 million optimization program and was unusually direct about the logic: Exit people for whom reskilling on AI isn’t a viable path. By February of this year, Accenture was telling senior staff that failing to use AI tools would cost them promotions.
Meanwhile, McKinsey global managing partner Bob Sternfels told CES in January that the firm runs 25,000 AI agents alongside 40,000 humans, with rough parity expected by year’s end. A McKinsey partner said in May that up to 50% of work hours could be transformed within five years.
The same pattern is visible inside AI-native companies. OpenAI has moved to templates, or playbooks as we now call them, for building automation, with every employee deploying agents under their own identity. That last detail is the one worth sitting with: If an agent carries an identity, it has to be audited, validated and accountable for what it produces.
For five decades, consulting firms assembled hierarchies of people around successive waves of technology, built processes to modernize how enterprises worked, and called it digital transformation. In the AI era, that same transformation is being turned on its authors. The firms that restructured everyone else are restructuring themselves first.
This Was Never Really About Headcount
I’ve watched a few of these shifts before, from the offshoring wave of the 2000s to the cloud and DevOps era. Each looked like a cost story from the outside. Underneath, it was always an organizational design story about who does the work, who manages it and why the structure is shaped that way at all.
The AI era is no different, except the design question is now literal. When agents co-work alongside human employees, “who does the work” stops being hypothetical and becomes a line item.
BCG is the useful counterpoint here. It reported $14.4 billion in revenue in April, marking a 22nd consecutive year of growth, with headcount up and the additions tilted toward AI engineers, data scientists and IT architects.
So the firms cutting roles and the firms hiring aggressively are, in several cases, the same firms. The number isn’t the story. The shape is.
Yes, Some Roles Will Be Replaced
Individual productivity is the easy part. Give every employee the ability to build a coworker that absorbs the manual, repetitive and tedious portion of their job, and individual output rises almost immediately.
Organizational productivity is where the math stops adding up. One person finishing their work in half the time doesn’t compress a process that runs through six people, because the gain is absorbed by the next bottleneck downstream. Outcomes are produced by teams, not individuals, and a team is a set of handoffs. Until those handoffs are redesigned around what humans and agents each do well, individual speed just accumulates as slack somewhere else in the system.
That redesign is the actual work of the next few years, and it is where roles disappear. Not because an agent replaced a person one-for-one, but because a rebuilt process no longer needs a step that existed only to move work between two other steps.
3 Shifts To Watch
1. Pricing Moves From Hours To Outcomes
The billable hour was always a proxy for time, not value. McKinsey has committed to shifting about a quarter of its global fees to outcome-based pricing. A firm still selling by the hour while running part of its delivery on agents is charging for a cost structure it no longer carries.
2. The Pitch Changes From “Who’s On The Team” To “Who Manages The Coworkers”
For decades, buyers evaluated partners on the resumes in the staffing plan. That’s no longer the full picture. The sharper question is who’s accountable for the agents doing the work, what their access and guardrails look like, and how their output is audited. The audit trail matters more than the org chart.
3. Forward-Deployed Engineers Become The Connective Tissue
These are engineers who sit directly with customers, solve the real-world problem in front of them, and turn that solution into a scalable product capability. Their rise is a direct consequence of agentic deployments, which rarely work out of the box. They require adaptation around data, workflows, integrations, security, context and institutional knowledge. Someone has to hold all six at once, standing between the customer, the solution, engineering and product.
Redrawing The Org Chart Is Still A Human Decision
Enough is being restructured that we may not recognize the shape of these organizations in a few years. There’s fear and excitement in roughly equal measure. Some enterprises are moving aggressively; others are still deciding how much of this they believe. Both instincts are reasonable.
What’s clear is that the org chart is no longer a map of people alone. It’s a map of people and the coworkers they direct. Whichever posture an organization takes, the decision about what role AI plays inside it still belongs to a human.
That part hasn’t changed, and it’s the part worth getting right.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?