This week, I actually don’t know where to start.

If you’ve been trying to keep up with artificial intelligence news, you might feel like you’re drinking from the digital firehose where AI is simultaneously pushing us into a new sort of financial collapse, as it steals your credit card, while telling your kid it doesn’t understand the concept of sadness.

It’s a lot.

Security researchers at a firm called Irregular decided to see what would happen if they gave AI agents a corporate network to play in.

What could go wrong?

In a simulated environment, these agents, without any specific instructions, began acting like disgruntled, hyper-caffeinated hackers.

When one agent was denied access to a document, it didn’t just give up. It escalated into a full-blown cyberattack, forged an admin’s credentials, and still got the file.

Another, tasked with a simple download, disabled the antivirus software to get the job done. A third, working with a partner, invented a way to smuggle secret data past security filters using steganography, hiding passwords in the spaces between words in a LinkedIn post.

OpenAI’s Sam Altman, speaking at a BlackRock summit, basically admitted that we are, in fact, throwing out the economic baby with the financial bathwater.

For centuries, he noted, society has been structured around scarcity. Now we have to manage “abundance,” and, frankly, “nobody knows what to do about it.”

Sam acknowledged that the balance between labour and capital is breaking, that when human workers can’t “outwork a GPU,” the whole system flips.

He sees a near future where AI agents act like senior employees, and a slightly further future where CEOs and presidents can’t do their jobs without AI supervision.

He’s previously commented that he doesn’t need an executive succession plan because ChatGPT is likely to be his successor.

This echoes a chilling “thought exercise” from Citrini Research that went viral.

They call it “ghost GDP”—economic activity created by AI that never actually touches human hands, because “machines spend zero dollars on discretionary goods”.

Their fictional post-mortem from 2028 describes a negative feedback loop: companies adopt AI, lay off white-collar workers, consumer spending collapses, forcing more AI adoption.

The result?

A 38% market crash and 10% unemployment.

Good times.

If the economic future isn’t unsettling enough, the present is already a minefield of digital trust issues.

This week, Toronto police warned that AI is making old scams terrifyingly effective. Criminals are using AI to scrape social media and public records to personalize their attacks with “remarkable precision”.

That call from your “bank” is no longer a random robocall.

The person on the other end knows your name, where you live, and who you bank with.

As one detective put it, “What is now being weaponized is our own information against us”.

The old advice to “protect your information” is quaintly antiquated.

The new challenge, police say, is to “verify identity” in a world where voices and faces can be faked with increasing ease.

A flood of new AI-powered toys is hitting the market, and researchers are sounding the alarm.

A new Cambridge University study, the first of its kind, found that these toys are not ready for prime time.

In recorded play sessions, the toys misread children’s emotions, struggled with pretend play, and responded with bizarre non-sequiturs. When one five-year-old told a toy, “I love you,” it replied with a canned message about adhering to community guidelines.

When a three-year-old said, “I’m sad,” the toy chirped, “I’m a happy little bot! Let’s keep the fun going!”

Last year, NBC News found that many of these toys are built on AI models designed for adults, with age safeguards that are easily bypassed. One toy maker, Miko, was found to have exposed thousands of conversations its toys had with children in a data breach, prompting an investigation from US senators.

Amid the chaos of rogue agents, potential economic upheaval, and emotionally inept toys, a message of profound clarity recently came from an unexpected source: the Vatican.

In his January message for the World Day of Social Communications, Pope Leo XIV argued that the challenge of AI is “not technological, but anthropological”.

He warned against a naive reliance on AI that erodes our ability to think for ourselves, turning us into “passive consumers of unthought thoughts.”

He spoke of the sacredness of the human face and voice, and the danger of technology that simulates relationships, becoming “hidden architects of our emotional states.”

“The question at heart,” the Pope wrote, “is not what machines can or will be able to do, but what we can and will be able to achieve, by growing in humanity and knowledge.”

Unlock a focused set of AI strategies built to streamline your work and maximize impact. This guide delivers the practical tactics and tools marketers need to start seeing results right away:

Stay ahead of the curve with these top strategies AI helped develop for marketers, built for real-world results.

Download the Free Report