ForbesPro
On a winter evening, a traveler watching airfares to New York refreshed the same page for the fifth time, only to find the numbers unchanged. A few hours later, while that traveler slept, the price quietly slipped, a hotel near Times Square launched a flash sale, and a pair of noise‑canceling headphones dipped into a short‑lived discount. The only entity awake for all three events was not a human bargain hunter. Still, an AI shopping agent working in the browser, quietly orchestrating the moment when those separate drops aligned.
That is the scenario Karma is betting on: a future in which shoppers set intentions, may it be a much-needed trip, a new laptop, or a wardrobe refresh, and then step away, leaving an AI shopping agent to track the market, compare options, and trigger the buy at the right time. In the crowded, margin‑tight world of e-commerce, that shift from manual searching to agentic AI is starting to look less like a novelty and more like an emerging norm.
E-commerce’s Turning Point
Analysts now forecast that the agentic AI market in retail and e-commerce will grow from approximately $46.7 billion in 2025 to roughly $175 billion by 2030, reflecting a compound annual growth rate of more than 30%. Separate estimates suggest that “agentic commerce” could account for 10% to 25% of U.S. ecommerce spending by the end of the decade. Against that backdrop, the question is no longer whether automation will impact shopping, but who will control the interface between the consumer and the marketplace.
Karma’s answer is a cross‑platform, easy-to-use, free AI shopping agent that runs through a desktop extension, Safari web extension, web dashboard, and mobile apps, following users wherever they browse. Instead of waiting at checkout, it embeds itself earlier in the journey, turning every product page, whether for flights, hotels, tech, or general marketplaces, into a signal that the system can watch over time.
Inside the Agentic AI Layer
Karma’s core promise is to move beyond one‑off coupon tools into what its team describes as an agentic AI layer for shopping. Users save items they are interested in; the system then tracks price drops, stock changes, and cross-site alternatives, while automatically testing coupon codes at checkout and surfacing better offers when they become available. Over time, those individual actions compound into a dataset. The company has reported several milestones, including 7 million users worldwide and millions of purchases made through the platform, which deliver meaningful savings to shoppers.
From the outside, the experience feels simple: a notification, a recommendation, a quietly applied promo code. Underneath, however, the agent operates continuously across verticals such as travel, tech, and fashion, where pricing rhythms differ, and timing is crucial. “What most people do not see is the volume of decisions,” says CEO Jonathan Friedman. “On a typical day, our AI shopping agent evaluates more possible combinations of prices, stock, and coupons than a human could examine in a year. The surprise is how often the ‘perfect moment’ looks accidental from the user’s point of view.”
Competing in a Crowded Field
Karma operates in a space defined by well‑known names such as Capital One Shopping, Coupert, and SlickDeals, all competing to surface savings in the browser. The company’s differentiation rests less on any single feature than on the insistence that the agent should act proactively across time, not just react at checkout. Instead of asking users to repeat the same searches, Karma’s agent is designed to internalize intent and carry the process forward, even when the shopper has closed the tab.
“The real story is not that an AI shopping agent can save you money,” CEO Jonathan Friedman says. “It is that, by 2030, delegating these choices may feel so natural that we will barely remember how exhausting it once was to chase every deal ourselves.” That remark echoes forecasts that by the end of the decade, AI agents will mediate a substantial share of online spending, turning what once felt like active hunting into largely delegated, background activity.
ForbesPro is Forbes Israel’s Promotional Paid Content Brand
