When AI Starts Deciding For You
19 Dec 2025 7:10 AM IST
Consider this. It’s Dec 2026. You have an AI agent.
You ask it to plan a trip to Bali. Flights booked. Resort selected.
Except the AI hallucinates. Instead of Bali, it books your stay in Pattaya, Thailand.
The money’s gone. And no human ever approved the decision.
Generative AI already helps us write, plan, and research.
Experts say the next phase is agentic AI, meaning systems that don’t just assist us, but act on our behalf.
Agentic AI means more convenience. And less control.
In the latest episode of The Signal Brief, we break down what agentic AI is and what its rise means for consumers.
Podcast mentioned: Shell Game, hosted by Evan Ratliff
The Core produces The Signal Brief. Follow us wherever you get your favourite podcasts. To check out the rest of our work, go to www.thecore.in
NOTE: A machine transcribed this episode. A human has looked at this text but there might still be errors. Please refer to the audio above, if you need to clarify something. If you want to give us feedback, please write to us at feedback@thecore.in.
—
TRANSCRIPT
Kudrat (Host): The first thing I do when I wake up, and I’m not proud to admit this, is that I check my phone.
I unlock it using facial recognition. I try not to open social media, so instead of Twitter or Instagram, I open the YouTube app. There are dozens of recommendations waiting for me, based on what YouTube thinks I’ll watch next.
I usually listen to a five-minute meditation. Then I catch up on the news.
On some days, I get a notification from Uber. “Lighter than usual traffic right now,” it says. “Book your cab.”
In just the first hour of my day, like so many of us, I’ve already interacted with artificial intelligence several times.
And none of this looks like what we usually think of as AI. Not ChatGPT, or Perplexity, or Claude–tools that have also become second nature to our professional and personal lives.
Kudrat (Host): My name is Kudrat Wadhwa and you’re listening to The Signal Brief. We don’t do hot takes. Instead, we bring you deep dives into the how and why of consumer trends.
In this episode, we take stock of where we are right now in the AI revolution, and what changes next, as we head into 2026.
Kudrat (Host): Many of us trace the AI moment to 2022, when OpenAI released ChatGPT to the public.
The platform could do so much. Make diet plans, write essays, even help you decode your crush’s texts.
But AI has been around for much longer.
At its core, artificial intelligence refers to computer systems that can learn from data, recognise patterns, and make predictions or decisions, often faster and at a much larger scale than humans.
To understand this better, The Signal Brief spoke to Vaibhav Jain, who runs an IT consultancy and works with clients from around the world.
Vaibhav: ChatGPT is not the inception of AI. It's a misconception that people generally have. The reason being that ChatGPT did make AI visible and conversational. It actually put a face to technology. But AI has been so much woven into our daily lives already that we don't even realize that we are using it.
Kudrat (Host): In other words, we were using AI long before generative AI.
Traditional AI, or ambient AI, works quietly in the background: spotting fraud, filtering spam, recommending content.
Generative AI is different. It creates. Text, images, audio, even code. And for the first time, people who don’t know how to code can use these systems directly.
GenAI is everywhere. For instance, I use ChatGPT to plan travel itineraries or to quickly research a topic. AI is also deeply embedded in institutions, from banks to hospitals.
Vaibhav: Banks possibly have been using AI to detect fraudulent transactions since the 2000s. In healthcare, what is picking up really fast now is the diagnostics field.
AI has been reading X-rays and MRIs, and the reports that we are seeing are much more detailed because they can actually evaluate thousands of X-rays and thousands of MRIs very quickly. If you see weather reports, et cetera, farmers have also, if I talk about some developed countries, already been using it to monitor crop health and optimize irrigation according to that.
So photo apps, photographers use AI for photo editing, musicians for mastering tracks. Actually, it is endless. The list is endless.
Kudrat (Host): Though AI is everywhere, its form is changing.
Until now, most consumer AI has waited for instructions. You type a prompt. You click a button.
Experts say that the next wave in the AI revolution is Agentic AI.
Clips: Introduction to Agentic AI by SAS Software
Agentic AI Explained, Jensen Huang
Kudrat (Host): Until now, Gen AI could help you plan a trip. But, an agentic AI can also scan your calendar, compare flight prices, book tickets, reserve hotels, and remind you when to leave, all on its own.
Agentic AI won’t become mainstream overnight. But experts say that within the next 10 to 12 months, we’ll start seeing it embedded across many everyday tools.
There’s no doubt about its rise though, since Agentic AI offers several benefits to businesses, including Vaibhav’s own IT consultancy.
Vaibhav: So say, for example, in IT and software development, we have already started using AI to actually create the tasks for the team. We are focusing on how and what tasks are to be done for a software. And then the AI itself, according to the interaction we are having with it, is actually setting up all the tasks, creating the complete workflow that we have already communicated.
Kudrat (Host): Agentic AI promises convenience both to consumers and to businesses.
But the risks we’ve seen with generative AI don’t disappear. In fact, they get amplified.
Here’s Amit Prabhu, who runs Digiculum, an AI reskilling company, talking about this:
Amit: The other risk is the accuracy of the data, as we know, because AI systems are not always accurate and they are prone to hallucinations. So that is something that a consumer should be really aware of.
Kudrat (Host): Consider this. Say you have an ‘AI agent’ to help you plan your itinerary for a trip to Indonesia.
The AI agent plans the trip – finds the cheapest flights from Delhi to Bali, and comes up with seaside resort recs. When it comes time to booking, however, it hallucinates. Instead of booking your resort in Bali, the agent books one in Pattaya in Thailand instead. There goes all your hard-earned money.
An American journalist called Evan Ratliff ran an experiment with agentic AI. He started a company staffed only with himself and AI agents. Later, produced a podcast based on his experience. Here’s an excerpt from one episode, I’ll link the episode in the show notes below too.
This particular part shows Evan’s AI agents planning a weekend hike as a team bonding exercise.
CLIP: Episode 2: Outdoor Debugging, Shell Game
Kudrat (Host): And here’s another clip of Evan getting frustrated with his agents, because they just can’t stop talking over each other.
CLIP: Episode 2: Outdoor Debugging, Shell Game
Kudrat (Host): What this shows is that AI isn’t evil. Rather, it can be messy.
These systems are fast, confident, and autonomous, yes, but not always reliable.
This is the trade-off consumers are walking into as AI becomes more agentic.
The upside is convenience. Fewer clicks. Fewer decisions. Less friction.
The downside is control.
When systems act on our behalf, errors stop being small annoyances. They become missed flights, wrong purchases, locked accounts, or money that’s hard to recover.
And unlike a human mistake, it’s not always clear who is accountable when an AI gets it wrong.
This is why regulators are starting to insist on human oversight. The EU’s Artificial Intelligence Act mandates that companies design high-risk AI systems so that human operators can monitor, intervene, or even stop automated decisions when necessary.
And back home, in India, the Reserve Bank has set up expert committees and issued warnings about AI risk in banking, stressing the need for governance frameworks so automation doesn’t undercut consumer protection or people’s financial stability.
Kudrat (Host): AI didn’t arrive all at once. It slipped into our phones, our feeds, and now, potentially, our decisions.
As agentic AI becomes more common in 2026, consumers will have to decide how much autonomy they’re comfortable giving up for convenience.
Because the more these systems act on our behalf, the harder it becomes to pause them, question them, or undo their mistakes.
Outro: That's all for today. You just heard The Signal Brief. We don't do hot takes. Instead, we bring you deep dives into the how and why of consumer trends. The Core produces The Signal Brief. Follow us wherever you get your favourite podcasts.
To check out the rest of our work, go to www.thecore.in.
If you have feedback, we'd love to hear from you. Write to us at feedback@thecore.in or you can write to me personally at kudrat@thecore.in.
Thank you for listening.
