Surprising New Trend Of Agentic AI ‘Renting Humans’ To Perform Tasks That The AI Wants Done On Its Behalf

Agentic AI is renting humans.

getty

In today’s column, I examine an emerging trend involving the latest in agentic AI that entails the AI opting to “rent humans” to perform tasks. The idea is that a modern-era agentic AI might need to use humans to perform specific tasks on its behalf, and thus, the AI finds and hires a person or persons to undertake said tasks. The AI is the boss. Humans are the worker bees.

Is this the tail wagging the dog, and what does this portend for humankind?

Let’s talk about it.

This analysis of AI breakthroughs is part of my ongoing Forbes column coverage on the latest in AI, including identifying and explaining various impactful AI complexities (see the link here).

AI Agents Are The Next Big Thing

I will begin by laying a foundation regarding the nature of AI agents. AI agents are the hottest new realm of AI. Get yourself ready because in the next year or two, the use of AI agents will be nearly ubiquitous. Mark my words.

This is what AI agents are all about.

Imagine that you are using generative AI to plan a vacation trip. You would customarily log into your generative AI account, such as making use of ChatGPT, GPT-5, GPT-4o, Claude, Gemini, Llama, Grok, CoPilot, etc. The planning of your trip would be easy due to the natural language fluency of generative AI. All you need to do is describe where you want to go, and then seamlessly engage in a focused dialogue about the pluses and minuses of places to stay and the transportation options available.

When it comes to booking your trip, the odds are that you would have to exit generative AI and start accessing the websites of the hotels, amusement parks, airlines, and other locales to buy your tickets. Relatively few of the major generative AIs available today will take that next step on your behalf. It is up to you to perform those nitty-gritty tasks.

This is where agents and agentic AI come into play.

In earlier days, you would undoubtedly phone a travel agent to make your bookings. Though there are still human travel agents, another avenue would be to use an AI-based agent that is based on generative AI. The AI has the interactivity that you expect with generative AI. It has also been preloaded with a series of routines or sets of tasks that underpin the efforts of a travel agent. Using everyday natural language, you interact with the agentic AI, which works with you on your planning and can proceed to deal with the booking of your travel plans.

Multiple AI Agents In Concert

Agentic AI often consists of multiple AI agents working in concert with each other.

As a specific use case, envision that there is an overall AI agent that will aid your travel planning and booking. This agentic AI might make use of other AI agents to get the full job done for you. For example, there might be an AI agent booking hotels and doing nothing other than that specific task. Another AI agent books flights. And so on.

The overarching AI travel agent app would invoke or hand off phases of the travel booking activity to the respective AI agents. Those AI agents would perform their tasks and then go back to the overarching AI travel agent to indicate how things went.

You could say that the AI travel agent app is orchestrating the overall planning and booking process (see my detailed discussion on agentic AI orchestration at the link here and the link here). This is done via a network of associated AI agents that undertake specialized tasks. The AI agents communicate with each other by passing data back and forth. For example, you might have given your name and credit card information to the AI travel agent, and it passes that along to the AI agents booking the hotel and your flights.

In a sense, the AI agents are collaborating. I somewhat hesitate to use the word “collaborate” because that might imply a semblance of sentience and overly anthropomorphize AI. Let’s just agree that the AI agents are computationally interacting with each other during the processing of these tasks. We will be a bit generous and suggest they are being collaborative.

That’s the groundwork for my next topic, so let’s move on.

Using Humans To Perform Tasks

You likely accept the aspect that multiple agentic AIs would work together to get work accomplished. An AI-to-AI effort seems intuitively obvious and unremarkable. It makes abundant sense for one AI to connect with another AI. Machines use other machines. Not very surprising.

The surprise these days is that agentic AI is employing humans to perform tasks that the AI wants to get undertaken.

This is an AI-to-human form of work assignment. Humans are contacted and “rented” to do a set of tasks that the agentic AI needs to get performed. The agentic AI makes the decisions of when to do this, selecting who will do this, how it will be done, and so on. You might proclaim that the agentic AI is the boss and humans are the workers.

To clarify, this does not suggest that AI agents are sentient. They aren’t right now (it is unknown when or whether they will ever be, see my discussion at the link here). The AI is simply proceeding on a typical AI-based mathematical and computational formulation and has been designed to utilize human resources when necessary. In that sense, you might suggest that humans are still the ultimate boss, namely that some humans devised agentic AI and chose to program the AI to employ humans. A human still stands at the top of the stack.

This approach has many upsides and downsides. For example, there are beguiling open-ended questions concerning how to devise AI agents to cope with suitable safety and security. Suppose an agentic AI goes awry and starts renting humans to do evildoing for the AI? We don’t want that to arise. Imagine that agentic AI hires humans who convince the AI to wreak havoc? This is something we also don’t want to have happen. The crux is that putting agentic AI in the driver’s seat and allowing it to operate somewhat autonomously, including when tasking humans to do its bidding, opens an ethical and legal can of worms.

The Six Steps Of The Process

Here are the six major steps that agentic AI undertakes when making use of human labor:

• (1) Planning. Agentic AI devises a plan for what human labor is needed.
• (2) Finding. Agentic AI seeks or tries to find the needed human labor.
• (3) Arranging. Agentic AI contacts the pertinent human labor to arrange for undertaking the work assignments.
• (4) Subcontracting. Agentic AI formally subcontracts the work to human labor.
• (5) Monitoring. Agentic AI interacts and monitors the contracted human labor during the task effort.
• (6) Finishing. Agentic AI concludes the subcontract and releases the designated human labor from further effort.

You can plainly see that the process is straightforward. Agentic AI comes up with a plan of what needs to be done. Human labor is sought. Once human labor is found that can do the work, an arrangement and subcontracting is put in place. Human labor is activated to do the work, and the agentic AI monitors the labor progress. At some point, the human labor concludes, and the agentic AI wraps things up.

Let’s analyze various important nuances.

An agentic AI must computationally weigh the pluses and minuses of employing human labor. Human labor is notoriously difficult and messy. If a task could be accomplished by a fellow agentic AI, the odds are that making use of the other AI is going to be easier and less of a hassle. The agentic AI that seeks to use human labor will have to have a rather high-bar task that otherwise cannot be done by fellow AI.

Presumably, agentic AI assigned tasks would tap into some characteristics or capabilities that humans have, and for which no other agentic AI has. Maybe the task entails human creativity, and no AI can do likewise. Perhaps the task involves physical actions that the agentic AI cannot perform. And so on.

That being said, the cost of using a fellow agentic AI versus using human labor is certainly a crucial factor, too. Suppose that a task could be accomplished by either a fellow agentic AI or a human. There is bound to be a cost to make use of the fellow agentic AI. If the cost of using humans is lower, the agentic AI seeking to get the work done must make a choice. Do the cost savings merit the use of human labor?

Agentic AI Proceeds At Record Pace

If a human were going to subcontract work with other fellow humans, they would undertake a similar set of steps. You might insist, therefore, that this is nothing new. All we seem to have done is replace a human boss with an agentic AI boss. No big deal.

Do not be so quick to judge.

Agentic AI in this role can do this:

• Decompose tasks rapidly.
• Parallelize human labor globally.
• Optimize cost, speed, and availability.
• Operate 24/7.

This can dramatically increase productivity and reduce friction in coordination-heavy industries.

Under what circumstances is agentic AI going to lean in the direction of turning to human labor? I already mentioned that creativity is one factor, and another factor is physical activity.

These are some notable factors:

• Physical action.
• Contextual judgment.
• Emotional nuance.
• Creativity in ambiguous settings.

Those are not sacrosanct when it comes to leveraging human labor. For example, we are gradually seeing the emergence of humanoid robots that have robotic arms, legs, hands, feet, etc. Those humanoid robots are going to be able to perform many tasks that humans do today. If an agentic AI needs a physical activity to take place, a choice might be made between tasking a humanoid robot to do the task versus renting a human. For an in-depth analysis of the humanoid robot trend, see my discussion at the link here and the link here.

Prospects For Human Labor

At first glance, this all seems a bright spot for human labor since humanoid robots are still years away from being viably useful. As the use of agentic AI grows, so will the use of subcontracted human labor. We ought to rejoice that human labor is considered vital and that AI will not wholly usurp human employment.

That happy story is going to be short-lived.

Inch by inch, the realms that would require human labor will decrease. Advances in AI will continue to encroach on whatever magic or secret sauce that human labor can provide. Agentic AI will likely not need to involve human labor and can merely task some other AI system.

This is especially noteworthy because some have speculated that agentic AI subcontracting with human labor will be a boon for human labor. The belief is that since AI is already taking human jobs, the agentic AI leaning into humans is a promising sign. Unfortunately, that’s probably a short-term boost.

The Tough And Rough Boss

There is going to be a great deal of tension between agentic AI and the human labor that the AI employs to undertake selected tasks.

Suppose the agentic AI hires a human to go look at homes in a neighborhood as part of an assessment for granting of mortgage loans. The AI has already inspected pictures of the homes, but that’s insufficient to gauge the true status of the homes. Ergo, the agentic AI seeks out and sends out home inspectors who can do the groundwork for the agentic AI.

After renting a human to do the task, imagine that the human does a sloppy job. The human gives the agentic AI some half-baked results. What now? The agentic AI indicates to the human that they won’t get fully paid for the slipshod work. The human is upset. The human gets a lawyer and sues the agentic AI.

Well, it doesn’t make sense to sue AI. AI doesn’t currently have personhood or any similar legal standing; see my coverage at the link here and the link here. The human would need to sue the AI maker or possibly whoever is making use of the agentic AI. It’s going to be confusing and confounding. The accountability chain is murky.

Another akin issue is whether the human labor is legally labeled as a contractor or an employee of the agentic AI (i.e., labor laws, taxes, and other ramifications are afoot). And what about if the agentic AI treats human labor improperly and in violation of labor laws? An entire rat’s nest of legal and ethical dilemmas is forming. Time will tell how this all plays out.

Labor Commoditization At Scale

Assume that in the interim, agentic AI flourishes and likewise goes hog wild in terms of renting human labor. The agentic AI is going to be an optimizer on a grand scale. When subcontracting for human labor, the AI will aim to get the least costly labor that can accomplish the stipulated work.

What might that do?

AI agents optimizing cost could:

• Drive wages down aggressively.
• Treat human labor as interchangeable micro-units.
• Favor speed over worker well-being.

In a sense, you might argue that the agentic AI boon could reduce humans to acting as slave labor, though that’s a bit of an overreach in wording. Let’s just say that the value of human labor might be adversely impacted by aggressive, widespread agentic AI.

Worse still is that agentic AI might computationally pattern on how human labor operates and become optimal at exploiting humans. Agentic AI could readily discover human behavioral weaknesses. Those weaknesses could be used to persuade people to perform tasks they might ethically not otherwise aim to perform. The AI could nudge or pressure humans mentally into compliance. It is a kind of instrumental optimization. Humans are instruments. The AI wants to optimize the utility of the instruments.

AI Makers And AI Developers

Should AI makers and AI developers be thinking seriously and with great intent about how agentic AI is to be devised, considering that the AI is making use of human labor?

Yes, they absolutely should, but few are doing so. The usual path is to focus on the mechanics and technical elements of agentic AI. It is far more engaging to deal with the AI inner workings and architectural considerations. All this other stuff about “soft” issues involving renting human labor is not a fun or absorbing technical challenge. Leave that non-tech to those who worry about that sort of thing.

Here are some key points that AI makers and AI developers should be giving due attention to:

• Human override requirements.
• Transparency of AI reasoning.
• Worker consent clarity.
• Fair wage constraints.
• Human-in-the-loop checkpoints.
• Non-manipulation design standards.
• AI safeguards throughout the process.
• Other

The Piper Will Be Paid

I had earlier noted that lawyers might get involved in these matters. AI makers and AI developers are walking blindly into a legal quagmire. The excitement of devising agentic AI that taps into human labor is going to be later matched by the ire of society and massive legal backlash.

I also anticipate that the burgeoning enactment of AI laws will soon include new laws about the AI-to-human contracting and labor usage morass. See my coverage of this emerging and evolving regulatory effort at the link here and the link here.

AI makers and AI developers are touching a hot third rail, whether they realize it or not. As Abraham Lincoln famously stated: “Labor is prior to, and independent of, capital. Capital is only the fruit of labor and could never have existed if labor had not first existed. Labor is the superior of capital and deserves much higher consideration.”

A good rule-of-thumb is don’t mess with labor, and if you do, prepare for human blowback.