Samsung’s Shift to ‘Agentic AI’ Aims to Revive Flagship Appeal
The shift from standalone AI features to proactive, task-handling assistants aims to revive flagship appeal.
Samsung Electronics celebrated record fourth-quarter earnings last month, largely due to booming memory chip demand, but the picture was less triumphant for its smartphone division. The Mobile Experience, or MX, unit reported its weakest quarterly operating profit in three years under mounting pressure from soaring component costs.
Still, one line buried deep in Samsung’s earnings release stood out: “MX will expand sales centered on flagship products with the launch of Galaxy S26, and strengthen leadership in the AI smartphone market through an agentic AI experience.”
Though Samsung has not elaborated on what the S26’s agentic experience will look like, it was the first time it publicly referred to “agentic AI” in such explicit terms. It was an unusual choice of phrase in a financial disclosure, and one that industry watchers say carries strategic weight.
The Shift Behind the Phrase: What is Agentic AI?
“Samsung’s shift from ‘AI features’ to ‘agentic AI’ is not a marketing flourish,” Jan Lorbach, senior director of Strategic Insights at the German market research GfK, told The Korea Herald. “Consumers still don’t see AI as a primary reason to buy a phone. But the idea of an agent, something that helps with travel, shopping, or scheduling, feels like a solution to real problems.”
Samsung has aggressively marketed its Galaxy AI capabilities since early 2024, offering tools like real-time translation, AI photo editing, and Gemini-powered voice assistance. But ‘agentic AI’ suggests a more proactive, integrated layer of intelligence.
Lorbach sees the agentic AI framing as part of a necessary shift in a year when the average selling price of Samsung smartphones dropped 17 percent and production costs continued to rise.
“To justify higher prices, brands have to show added value. If the hardware cost curve is no longer favorable, that value has to come through software.”
Nabila Popal, Senior Research Director at IDC, also told The Korea Herald that Samsung’s agentic AI push comes at a critical moment for its smartphone business. “Flagship sales are under pressure, and the market is getting tougher,” she said.
“But Samsung has a chance to pull ahead. Especially if Apple continues to move slowly on AI. Right now, Samsung seems likely to launch agentic AI first, and that gives them a window to shape the narrative.”
What Gemini on Galaxy Can and Can’t Do
Samsung’s current AI experience, branded as Galaxy AI, has already evolved toward more integrated workflows.
Gemini Live, developed by Google and set as the default voice assistant on the Galaxy S25 series and other Galaxy flagships in 2025, can currently perform some multi-step tasks across supported apps. For instance, users can ask Gemini to look up an event, summarize it, and add it to their calendar, all in a single prompt.
Since January 2025, Gemini’s app extensions include Samsung Calendar, Notes, Reminder and Clock, in addition to Google services and a limited set of third-party apps such as WhatsApp and Spotify. However, the actual capabilities remain constrained. Calendar events can be created and edited, but not shared. Notes can be retrieved and summarized, but not tagged or collaborated on.
Samsung may be exploring ways to augment this experience further by upgrading its in-house voice assistant, Bixby. Once the primary assistant on Galaxy devices, Bixby has been relegated to a secondary role since Gemini’s introduction.
“There’s speculation that Samsung may be planning to enhance Bixby, possibly integrating technologies like Perplexity for better general knowledge queries ,” said an industry official who requested anonymity. “On top of that, Bixby could be upgraded to make proactive suggestions and handle more context-aware tasks, especially more on-device actions or across non-Google apps where Gemini can’t go.”
Trust as a Differentiator in the AI Landscape
But for such agentic systems to function meaningfully, they often need full access to personal user data. That makes brand trust and data protection central to the equation.
Professor Choi Byung-ho, a research lead at Korea University’s Human-Inspired AI Institute, said, “You have to be willing to fully delegate. And that only happens when there’s institutional and technical trust,” he said. “Samsung, as a device brand with security baked in at the hardware level, is in a position to make that promise credibly.”
“This is where Samsung has a real structural edge. Samsung’s Knox and on-device privacy architecture are already in place. As the industry moves beyond chatbots, the brand that can say ‘your data never leaves the device’ holds the high ground.”
He added that AI features today still require too much manual input. “People say AI is impressive, but don’t use it regularly. That’s because the burden is still on the user. What makes an agent valuable is that it anticipates routine tasks and quietly handles them.”
“Regardless of who wins, the reality is AI still isn’t driving upgrades,” Popal from IDC said. “Consumers remain skeptical. But Samsung is one of the few Android players with the portfolio depth and scale to keep iterating on this across many form factors and markets.”
