DoorDash stock drops more than 5% as AI ‘agent’ memo rattles delivery apps

February 23, 2026
DoorDash stock drops more than 5% as AI ‘agent’ memo rattles delivery apps

New York, Feb 23, 2026, 12:07 (EST) — Regular session

DoorDash Inc (DASH) dropped over 5% early Monday, caught in a fresh round of AI-driven jitters rattling richly valued consumer platforms. Shares traded at $166.79 by late morning, down 5.4% after dipping to $161.50 earlier. Uber Technologies gave up 3.7%, while Maplebear Inc, Instacart’s parent, shed 1.4%.

This shift is catching investors as they revisit a basic question: is DoorDash’s strength really baked into customer routines, or does it come down to logistics others can’t easily mimic? When nerves start jangling, companies that look like pure toll collectors tend to take the first hit.

The discussion intensified after Citrini Research and Alap Shah put out a “thought exercise” on Feb. 22, suggesting AI agents could one day eliminate fees from routine transactions. Their note labeled DoorDash as “the poster child,” pointing out that “habitual app loyalty … simply didn’t exist for a machine.” In their scenario, bots would scan all available delivery choices and go for the lowest price every time. Citrini Research

DoorDash shares slid, joining a wider market pullback as tariff worries resurfaced. “You simply can’t bet against Trump,” said Thomas Hayes, chairman at Great Hill Capital, after President Donald Trump rolled out a fresh 15% global duty. The move followed a Supreme Court decision scrapping earlier, broader tariffs. Reuters

DoorDash paused operations in Connecticut and several other states battered by the nor’easter, with blizzard conditions sweeping the area, the company confirmed.

Sellers are hitting a stock that not long ago picked up momentum on upbeat guidance. Last week, DoorDash projected first-quarter marketplace gross order value between $31 billion and $31.8 billion. The company also flagged plans to pour several hundred million dollars into its tech stack in 2026, aiming to bring DoorDash, Wolt, and Deliveroo under a single platform. That said, its adjusted EBITDA guidance—$675 million to $775 million—fell short of consensus. RBC Capital Markets analysts called out grocery and retail as bright spots, while Bernstein’s Nikhil Devnani argued DoorDash is simply doing what it does, with reinvestment baked into its culture. On valuation: DoorDash’s forward P/E sits near 51, compared to Instacart at about 15 and Uber at 21.

Now, traders are eyeing whether the AI-driven selloff spills into more fee-collecting platforms or just fizzles out as fast as it came. DoorDash’s habit of outspending rivals—while others ramp up promotions—remains under scrutiny.

U.S. delivery is still a battleground, and there’s more at stake now than just restaurant orders. New competitors or better price-comparison tools could force fees down, quickly eroding the platform’s “take rate” — that’s the cut they keep.

But a scenario memo is one thing—turning that into a credible earnings model is another. DoorDash might lean on AI to trim support expenses, make routing smarter, and hold on to orders, even as fresh competition enters the app space.

Nvidia’s results drop after the bell on Wednesday, Feb. 25, and there’s no overstating how much traders will be watching for any hint of “AI” momentum — it’s a high-stakes read on market risk appetite. DoorDash, meanwhile, could see its next move sparked by a fresh wave of analyst reactions or tariff chatter, both of which have a habit of rattling customers and investors alike. Nvidia

Technology News Today

  • UTS study links 3D airway model to personalised respiratory therapy
    April 13, 2026, 8:29 PM EDT. UTS researchers used a patient-specific, CT-derived 3D airway model to simulate continuous high-frequency oscillation therapy (CHFO) and map how pressure, wall shear stress and loading move through the conducting airways. Lead author Dr Suvash C. Saha of the UTS School of Mechanical and Mechatronic Engineering says CHFO supports airway clearance and lung expansion, but its effects vary by region. The study shows the throat and upper airway bear stronger pressure and friction, while larger upper-airway regions carry more of the force. Increasing pressure strengthens support but does not shift where effects concentrate. Airway anatomy largely fixes the loading hotspots, underscoring the need for patient-specific device settings and evidence-based design to improve safety, comfort and effectiveness. The work is published in Respiratory Physiology & Neurobiology.