McKinsey and LinkedIn’s 20 new agent AI job boxes are all coming soon (Photo by Smith Collection/Gado/Getty Images).
Getty Images
The main headline about Agentic AI and jobs is that it will kill jobs. Data tells a second story that gets less attention.
January 2026 on LinkedIn The numbers were assigned by analysis. AI has already added more than 1.3 million new roles, such as AI Engineer, Forward Deployment Engineer, and Data Annotator, as well as more than 600,000 AI-enabled data center jobs. LinkedIn calls this the New Collar Era, calling for a workforce that combines knowledge work, advanced technical skills, and obvious human strengths. AI engineers are currently ranked as the #1 fastest-growing job in the U.S. for the second year in a row.
Boxes added a dimension to the trend this week. respond to new york times In an article about AI-driven hiring, CEO Aaron Levy explained that positions that didn’t exist at the company two years ago are expanding across engineering, IT, and go-to-market. That means roles to evaluate new AI models, internal automation engineers to reshape how the company works, and professional services teams to perform forward-deployed engineering activities to help customers live faster.
What he meant was that Box was no exception. This is an early indication that the same changes are occurring within our customers.
Institutional research agrees.
world economic forum‘s Future of Jobs Report predicts 170 million new roles will be created by 2030, compared to 92 million unemployed people, resulting in a net increase of 78 million jobs, with AI and machine learning specialists among the world’s three fastest-growing occupations. McKinsey envisions its destination as an agent organization where generalists become orchestrators of AI agents, creating entirely new roles to oversee, guide, and govern AI agents.
In other words, the question is not whether AI will create jobs. It shows that the company has actually restructured, rather than just announcing a strategy, and it shows which jobs are which.
Here are the 20 most important.
20 Future Jobs for Agentic Ai in Tech and Beyond
sandy carter
AI and agent AI technical core
1. Forward Deployment Engineer
Embed it inside your customers and build and orchestrate agent workflows for your stack. Palantir created it. OpenAI, Anthropic, Google, and EY all employ this sector, and LinkedIn lists it as one of the three roles driving the 1.3 million number.
2. AI engineer
It is the fastest growing title in the country with 75,000 of the 639,000 new AI posts on LinkedIn. Build and run AI products, agents, and language models at scale. It is often renamed Applied AI, GenAI, or LLM Engineer.
3. AI evaluation engineer
Evaluate Box’s adoption of new AI models. Build test suites that catch hallucinations, regressions, and biases before they reach the customer. It is now a permanent discipline that changes every time the model is updated.
4. Context Engineer
From search pipelines to memory, design systems that feed your models with the right information at the right time. When an agent takes on a multi-step task, this role determines whether the agent is successful.
5. AI Agent Architect
Models coordinate tools and design autonomous multi-agent systems that chain complex actions. Engineering core for agent commerce, support, and operations.
6. Chief AI Officer
Responsible for company-wide AI strategy and ethical implementation. According to IBM research, one in four companies currently have it installed, and two-thirds expect most to do so within two years. Smaller companies hire partial CAIOs.
7. Head of AI Governance and Ethics
A growing number of new roles are around AI ethics and governance.
getty
Keep AI systems transparent and accountable, align with EU AI and state laws, and manage bias, impartiality, and privacy risks. It is already mandatory in regulated industries.
8. AI Enablement Lead
Support Box’s internal automation hiring. Run an adoption program, train employees, and track whether the tools are actually being used. McKinsey calls this agent coaching.
9. AI Red Team Engineer
Adversarially stress test your AI systems to investigate jailbreaks and dangerous behavior before deployment. The cost of skipping this increases as the agent gains authority to perform an action.
10.Decision Engineer
McKinsey’s T-shaped experts rethink workflows and protect quality. Designing for humans and AI to share the work, and what happens when the model becomes unreliable. It’s a natural axis for consultants and analysts.
Agentic AI’s role beyond engineering
There’s a bigger jobs story here. McKinsey put it bluntly: “The biggest new category is not engineers, but front-line workers in sales, service, human resources, and operations powered by AI.”
These 10 roles are more than technical roles, and most of them don’t require any code.
11. AI consultant and strategist
The second fastest-growing role on LinkedIn is behind AI Engineer and is much less technical. Translate features into business cases and determine deployment order. A direct path for consultants and business leaders.
12. AI Product Manager
Own the roadmap for your AI capabilities and agent products and sit between what your models can do and what the market will pay. As every product becomes an AI product, non-engineering roles will become some of the most sought after.
13. Agent Supervisor
M-shaped boss at McKinsey Tune the agent and a hybrid workforce. KPMG claims that agents belong to an organizational chart with reporting lines and are managed like human resources. The people running that digital team are managers, not engineers.
14. AI trainers and data annotators
AI teachers label and learn data models and adjust their output through human feedback. LinkedIn names data annotator among the three roles behind 1.3 million people. One of the most accessible entry points to AI jobs.
15. Conversation Designer
agent’s voice. AI shapes how you talk to customers and employees, its tone, flow, and fallbacks in times of uncertainty. It’s about writing and experience, not coding.
16. Human Agent Workforce Lead
New HR functions. KPMG argues that HR departments need to manage agent performance and engagement in the same way as human agents. Responsible for workforce planning that considers both talent and agents to reimagine career paths for hybrid teams.
17. AI Legal and Policy Advisor
We own our intellectual property, data rights, and maintain compliance with EU AI laws and a growing number of state laws. When an agent takes an action that costs money, the responsibility falls to a predetermined person.
18. AI Augmented Sales and GTM Specialist
McKinsey’s largest category: Front-line salespeople who spend less time in systems and more time with humans while agents handle research and follow-up. Rather than disappearing, roles change form.
19. AI Customer Success Manager
Box captures the demand described by customers. Drive adoption and prove results within your account, not just sales. We rebuilt customer success around agent deployment and expansion.
20. AI Auditor and Assurance Officer
Independently verify that your AI systems meet standards such as ISO 42001, the first international standard for AI management systems. A natural pivot for risk, audit and compliance professionals.
What does this actually tell us from agenttic AI?
Each of these 20 roles is evidence of companies changing not just their press releases but their organizational charts. They are only hired if management commits to running the business through an agent. That’s why KPMG found that 70% of large companies plan to reorganize within three years.
Ten of those jobs are not engineering jobs. These are strategic, product, human resources, legal, sales, and assurance roles that didn’t exist in their current form two years ago. Employment data shows that is becoming the operating model. Jobs eliminated by AI will continue to make headlines. The jobs created by Agentic AI will silently determine which companies will still be around in five years.
