Generative artificial intelligence (AI) has come to increasingly fit neatly into that lineage.

Findings from the PYMNTS Intelligence July 2025 PYMNTS Data Book, “The Two Faces of AI: Gen AI’s Triumph Meets Agentic AI’s Caution,” reveal that CFOs love generative AI. Nearly 9 in 10 report strong ROI from pilot deployments, and an overwhelming 98% say they’re comfortable using it to inform strategic planning.

Yet when the conversation shifts from copilots and dashboards to fully autonomous “agentic AI” systems, software that can act on instructions, make decisions, and execute workflows without human hand-holding, the enthusiasm from the finance function plummets. Just 15% of finance leaders are even considering deployment.

This trust gap is more than a cautious pause. It reveals a deeper tension in corporate DNA: between a legacy architecture designed to mitigate risk and a new generation of systems designed to act. Where generative AI has found traction in summarizing reports or accelerating analysis, agentic AI demands something CFOs are far less ready to give: permission to decide.

Why Agentic AI Feels Different

Generative AI won finance leaders over by making their lives easier without upending the rules. It accelerates analysis, drafts explanations, and surfaces hidden risks. It works inside existing processes and leaves final decisions to people.

That made the ROI for generative AI obvious: faster closes, better forecasts and teams that can do more with less. It’s the kind of technology finance chiefs have embraced for decades.

Agentic AI is different. These systems don’t just suggest — they act. They can reconcile accounts, process transactions or file compliance reports automatically. That autonomy is exactly what the PYMNTS Intelligence report found rattles finance chiefs. Executives who love Gen AI when it writes reports or crunches scenarios can slam on the brakes when agentic machines start to move money or approve deals.

Governance is the first worry. Who signs off when a machine moves money? Visibility is another. Once an AI agent logs into a system over encrypted channels, security teams may have no idea what it’s really doing. And accountability is the big one: if an autonomous system makes a mistake in a tax filing, no regulator will accept “the software decided” as an excuse.

Read the report: The Two Faces of AI: Gen AI’s Triumph Meets Agentic AI’s Caution

The black-box nature of AI doesn’t help. Unlike traditional scripts or rules engines, agentic systems use probabilistic reasoning. They don’t always produce a clear audit trail. For executives whose careers depend on being able to explain every number, that’s a deal breaker.

Legacy infrastructure makes things worse. Finance data is scattered across enterprise software, procurement platforms, and banking portals. To work autonomously, AI would need seamless access to all of them, which means threading through a maze of authentication systems and siloed permissions.

Enterprises already struggle to manage those identities for employees. Extending them to machines that act like employees, only faster and harder to monitor, could be a recipe for hesitation.

If autonomous systems are going to move beyond experiments, they’ll need to prove their value in hard numbers. Finance chiefs want to see cycle times shrink, errors fall, and working capital improve. They want audits to be faster, not messier.

The irony is that CFOs don’t need AI to be flawless. They need it to be explainable. In other words, transparency is the killer feature.

Unless agentic AI can show that kind of return, it may stay parked in the “idea” column instead of the project pipeline.

As more kids turn to artificial intelligence to answer questions or help them understand their homework, some appear to be forming too close a relationship with services such as ChatGPT — and that is taking a toll on their mental health.

“AI psychosis,” while not an official clinical diagnosis, is a term clinicians are using to describe children who appear to be forming emotional bonds with AI, according to Dr. Ashley Maxie-Moreman, clinical psychologist at Children’s National Hospital in D.C.

Maxie-Moreman said symptoms can include delusions of grandeur, paranoia, fantastical relationships with AI, and even detachment from reality.

“Especially teens and young adults are engaging with generative AI for excessive periods of time, and forming these sort of fantastical relationships with AI,” she said.

In addition to forming close bonds with AI, those struggling with paranoia may see their condition worsen, with AI potentially affirming paranoid beliefs.

“I think that’s more on the extreme end,” Maxie-Moreman said.

More commonly, she said, young people are turning to generative AI for emotional support. They are sharing information about their emotional well-being, such as feeling depressed, anxious, socially isolated or having suicidal thoughts. The responses they receive from AI vary.

“And I think on the more concerning end, generative AI, at times, has either encouraged youth to move forward with plans or has not connected them to the appropriate resources or flagged any crisis support,” Maxie-Moreman said.

“It almost feels like this is a burgeoning epidemic,” she added. “Just in the past couple of weeks, I’ve observed cases of this.”

Maxie-Moreman said kids who are already struggling with anxiety, depression, social isolation or academic stress are most at risk of developing these bonds with AI. That’s why, she said, if you suspect your child is suffering from those conditions, you should seek help.

“I think it’s really, really important to get your child connected to appropriate mental health services,” she said.

With AI psychosis, parents need to be on the lookout for symptoms. One could be a lack of desire to go to school.

“They’re coming up with a lot of excuses, like, ‘I’m feeling sick,’ or ‘I feel nauseous,’ and maybe you’re finding that the child is endorsing a lot of physical symptoms that are sometimes unfounded in relation to attending school,” Maxie-Moreman said.

Another sign is a child who appears to be isolating themselves and losing interest in things they used to look forward to, such as playing sports or hanging out with friends.

“I don’t want to be alarmist, but I do think it’s important for parents to be looking out for these things and to just have direct conversations with their kiddos,” she said.

Talking to a child about mental health concerns can be tricky, especially if they are teens who, as Maxie-Moreman noted, can be irritable and a bit moody. But having a conversation with them is key.

“I think not skirting around the bush is probably the most helpful thing. And I think teens tend to get a little bit annoyed with indirectness anyhow, so being direct is probably the best approach,” she said.

To help prevent these issues, Maxie-Moreman suggested parents start doing emotional check-ins with their children from a young age.

“Just making it sort of a norm in your household to have conversations about how your child is doing emotionally, checking in with them on a regular basis, is important. So starting at a young age is what I would recommend on the preventative end,” she said.

She also encouraged parents to talk to their children about the limits of the technology they use, including generative AI.

“I think that’s probably one of the biggest interventions that will be most helpful,” she said.

Maxie-Moreman said tech companies must also be held accountable.

“Ultimately, we have to hold our tech companies accountable, and they need to be implementing better safeguards, as opposed to just worrying about the commercialization of their products,” she said.

Get breaking news and daily headlines delivered to your email inbox by signing up here.

© 2025 WTOP. All Rights Reserved. This website is not intended for users located within the European Economic Area.

In today’s column, I examine an ongoing debate about who will have access to artificial general intelligence (AGI). AGI is purportedly on the horizon and will be AI so advanced that it acts intellectually on par with humans. The question arises as to whether everyone will be able to use AGI or whether only those who can afford to do so will have ready access.

Some ardently insist that if AGI is truly attained, it ought to be considered a worldwide public good, including that AGI would be freely available to all at any time and any place.

Let’s talk about it.

This analysis of an innovative AI breakthrough 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).

Heading Toward AGI And ASI

First, some fundamentals are required to set the stage for this weighty discussion.

There is a great deal of research going on to further advance AI. The general goal is to either reach artificial general intelligence (AGI) or maybe even the outstretched possibility of achieving artificial superintelligence (ASI).

AGI is AI that is considered on par with human intellect and can seemingly match our intelligence. ASI is AI that has gone beyond human intellect and would be superior in many if not all feasible ways. The idea is that ASI would be able to run circles around humans by outthinking us at every turn. For more details on the nature of conventional AI versus AGI and ASI, see my analysis at the link here.

We have not yet attained AGI.

In fact, it is unknown as to whether we will reach AGI, or that maybe AGI will be achievable in decades or perhaps centuries from now. The AGI attainment dates that are floating around are wildly varying and wildly unsubstantiated by any credible evidence or ironclad logic. ASI is even more beyond the pale when it comes to where we are currently with conventional AI.

Immense Capability Of AGI

Imagine how much you could accomplish if you had access to AGI.

AGI would be an amazing intellectual partner. Any decisions that you need to make can be bounced off AGI for feedback and added insights. AGI would be instrumental in your personal life. You might use AGI to teach you new skills or hone your own intellectual capacity. By and large, having AGI at your fingertips will be crucial to how your life proceeds.

The issue is this.

Suppose that AGI is provided by an AI maker that opts to charge for the use of the AGI. This seems like a reasonable approach by the AI maker since they undoubtedly want to recoup their investment made in devising AGI. Plus, they will have ongoing costs such as the running of AGI on expensive servers in costly data centers, along with doing upgrades and maintenance on AGI.

As they say, to the victor go the spoils.

Those who can afford to pay for AGI usage will have a leg up on everyone else. We will descend into a society of the AGI haves and the AGI have-nots. Things will get even worse since the AGI haves will presumably flourish in new ways and far outpace the AGI have-nots. Humans armed with AGI will excel while the rest of the populace is at a distinct disadvantage.

Some believe that if that’s how the post-AGI era plays out, something demonstrative would need to be done. We can’t simply allow the world to be divided into those that can afford AGI and those that cannot do so.

There must be a means to figure this out in some balancing way.

AGI As A Designated Public Good

One burgeoning idea is that AGI should be designated as an international public good.

If we reach AGI, no matter whether by an AI maker or some governmental entity or however reached, the AGI must be globally declared as a public good for all. There must be no barriers to the use of AGI. People everywhere are to have full and equal access to AGI. Period, end of story.

How might this be undertaken?

Some indicate that AGI should be handed over to the United Nations. The UN would be tasked with making sure that access to AGI was a worldwide facility. People in all nations and anywhere on the planet would readily be provided a login and full access to AGI. They could use AGI as much as they desired.

Furthermore, since there will undoubtedly be people who don’t know about AGI or don’t have equipment such as smartphones or Internet access, the UN would perform a global educational effort to get people up-to-speed. This would include providing equipment such as a smartphone and the network capacity needed to utilize AGI.

The AI maker that attained AGI would be given some compensation for their accomplishment but might be cut out of future compensation under the notion of a government taking of AGI from the AI maker. In an imminent domain type of action, their devised AGI would be taken from them and placed into a special infrastructure established to globally enable access to the AGI.

Lots of variations are being tossed around about whether the originating maker of AGI would be allowed to keep AGI and run it on behalf of the world or be set to the side as a result of the governmental taking of AGI. Pros and cons are hotly debated.

Worries About Human-AGI Evildoing

Wait for a second, some bellow vociferously. If everyone has ready access to AGI, at no cost and no barrier to usage, we are opening a Pandora’s box.

Consider this scenario. An evildoer opts to access AGI. They can do so without any cost concerns. So, they tell AGI to come up with a biochemical weapon. AGI runs and runs, inventing a new biochemical weapon that nobody knows about. The evildoer thanks AGI and proceeds to construct the weapon. They can use it to blackmail the world.

Not good.

One retort to this worry is that we would stridently instruct AGI to not undertake any effort that would be construed as detrimental to humankind. Thus, when someone tries to go down the evil route, AGI would stop them cold. The AGI would indicate that their request is not allowed.

But suppose an evildoer manages to trick AGI into aiding an evildoing plot. Perhaps the evildoer asks for something seemingly innocuous but is part of a subtle step in a larger evil plan. Step by step, they get AGI to give them solutions that can be ultimately pieced together into a larger evil purpose. The AGI was not the wiser for it.

It would be very challenging to somehow get AGI to in an ironclad fashion never reveal or respond in a manner that would completely avert all potential evildoing.

Registering People That Use AGI

A related facet is that some say we would want to make sure that the people using AGI are identifiable. In other words, even if ready access to all is provided, we should still require people to identify who they are. We cannot allow people to wantonly or anonymously use AGI.

Each person ought to be responsible for how they use AGI.

Whoa, comes the reply, if you require identification to use AGI, that’s an entirely different can of worms. People will naturally suspect that the AGI is going to monitor them. In turn, the AGI might tattle on them to government authorities. This is a dire pathway.

It would also be a somewhat impractical notion anyway. What kind of identification would be used? The identification could be faked. Some point out that with the latest high-tech, such as eyeball scans, presumably people could be distinctly and suitably identified.

Creating A Worldwide AGI Agency-Entity

There are those that express qualms about handing over AGI to the UN. The concern is that the UN might not be the most suitable choice to fully control and run AGI.

A suggested alternative is to establish a new entity that would oversee AGI. This would be a means of starting fresh. No prior baggage. The new agency entity would be formulated from the ground up as having the sole purpose of managing AGI.

Who would pay for this and how would the new entity be globally governed?

Various proposed approaches are being floated. Perhaps every country in the world would need to pay into a special fund for AGI. The monies would go to the ongoing costs of running AGI, along with administering the use of AGI. The new agency entity wouldn’t necessarily be expected to drive a profit and would be set up as a non-profit, perhaps as an NGO.

That sounds solid, but again there are those who doubt this new agency would be as beneficial and supportive as one might assume. For example, the agency might become corrupt and those running AGI start to use the AGI for their own nefarious purposes. The rest of the world might not know what’s happening.

A reply to this concern is that there would be auditors that routinely examine the new agency. They would report to the world at large. This would presumably keep the AGI-managing entity on the up and up.

Round and round these arguments go.

Universal Access To AGI

How do you feel about the claim that universal access to AGI is a must-do?

There are those who fall into the camp that AGI access must absolutely be a universal right. Nothing other than full universal access is to be undertaken. All humans deserve access to AGI.

Others say that AGI access is a privilege and isn’t necessarily going to be applicable to everyone. That’s the way the ball bounces. The world doesn’t owe everyone access to AGI.

Another variation is that we might have tiered access to AGI. Perhaps there would be AGI access at a minimum level for all, including associated constraints on usage, and then above that tier would be AGI usage with more open-ended facilities.

It’s a tough question.

At this time, it is a theoretical question since we don’t yet have AGI. That being said, there are predictions that we might have AGI in the next several years, perhaps by 2030 or so. The question is looming, and we probably should be figuring out what we are going to do when the moment arrives (assuming we do attain AGI).

Take a few reflective moments to ascertain where you stand on the thorny issue.

A final thought for now might be spurred by the famous line of Margaret Fuller, noted American journalist in the early 1800s: “If you have knowledge, let others light their candles in it.”