Tech earnings spotlight AI spending, privacy and enterprise control

StoneX sees combined 2027 free-cash losses exceed $309 billion across six hyperscalers, CoreWeave and Nebius

A technology-market roundup published Nov. 3, 2026, collected different views of where AI spending would pay off, who should control enterprise AI systems and how large AI operations would be financed. Apple reported fiscal fourth-quarter revenue and earnings above analyst expectations, and executives discussed privacy and costs during Monday’s earnings call. Palantir executives used an earnings call and shareholder letter to promote their platform’s role in enterprise AI, while separate analyst notes examined Tencent, Xiaomi and the hyperscaler market.

Apple’s revenue grew 12%. Tariffs added about 1 percentage point to that growth and 5 cents a share to earnings of $2.10. Apple said both figures would have exceeded analyst expectations even without the tariff benefit. Record-high sales in the company’s services business and a strong showing for the iPhone also supported the results.

During Apple’s Monday earnings call, executives mentioned privacy at least a dozen times, primarily while discussing Siri AI, the company’s new AI-driven personal agent. Chief Executive John Ternus said data Siri AI can access is protected, so not even Apple can access it. The company discussed rivals including Meta and Google while contrasting its approach with some peers that sell consumer data to third parties and advertisers.

“We just want to make sure that users are aware of what they’re giving access to,” Ternus said. “We want to make sure this is done in a way that protects security and privacy.”

Ternus also described the personal data on the iPhone as one of Apple’s biggest AI opportunities, pointing to contacts, communications, health data and financial data. Siri AI can access that information with permission and act as a personal assistant and AI agent, he said.

“So much of your personal data is on your iPhone, whether it’s your contacts, your communications, health data, financial data, all of that,” Ternus said. “We think that is a core differentiator that we’re excited about.”

Apple’s cost pressures moved in the opposite direction. Chief Financial Officer Kevan Parekh said higher memory prices were squeezing margins. Apple expects operating expenses in the current fiscal first quarter of $20.4 billion to $20.8 billion, ahead of the $20.26 billion analysts anticipated. Parekh attributed the higher costs mainly to rising memory prices and expenses associated with ramping Siri AI.

During the fourth quarter that had just ended, higher memory costs pressured the gross margin for both Apple overall and its products business. Tariff refunds kept both margins from being narrower, Parekh said.

Morningstar analysts said in a note that Tencent would benefit from greater user stickiness once it launches an AI agent in WeChat, strengthening the ecosystem behind a core business. The analysts identified three potential paths to direct monetization: taking a cut of agent-facilitated transactions; generating higher advertising revenue from better targeting, supported by richer user-intent signals gathered by the Xiaowei AI assistant; and offering a paid subscription tier for advanced AI features.

Morningstar said Tencent’s valuation did not require current AI spending to earn its cost of capital. The analysts also expected management to scale back AI spending if returns failed to meet targets.

Daiwa analyst John Choi said Xiaomi’s smartphone revenue likely strengthened in the third quarter as higher selling prices passed through, resilient demand persisted and new model launches supported sales. Daiwa now forecasts total quarterly revenue of 106 billion yuan.

Smartphone shipments likely fell, Choi said, because management applied strict volume control while shifting the product mix toward higher-end models. Average selling prices were expected to remain robust as Xiaomi continued its premiumization strategy. Daiwa raised its target price to HK$35.00 from HK$30.00 and maintained a buy rating; shares last traded at HK$28.64.

Deutsche Bank’s Bin Wang said Xiaomi’s car and AI business may have faced some sequential pressure in the third quarter. Deutsche Bank estimates revenue from the smart-EV, AI and other new-initiatives segment declined 4% on the quarter to 23.8 billion yuan, with vehicle deliveries flat.

Xiaomi has guided for 2026 corporate research and development expenses of 40 billion yuan, including a 16 billion yuan allocation to the smart-EV and AI segment. Wang said automotive research and development expenses may have dragged Xiaomi’s third-quarter adjusted net profit. Deutsche Bank kept a buy rating and a target price of HK$48.00; shares last traded at HK$28.74.

Palantir Chief Technology Officer Shyam Sankar challenged Anthropic ahead of the AI lab’s initial public offering this month, directing investors to a portion of Anthropic’s filing touting its work to improve AIG’s underwriting processes with AI.

“That entire workflow and the elegant application of AI software and humans runs in AIP and was implemented by Palantir and AIG,” Sankar said, referring to Palantir’s Artificial Intelligence Platform. “We are the diffusion layer, not just of choice, but of necessity.” The Wall Street Journal characterized Palantir as an outspoken critic of frontier AI labs. Palantir has positioned itself as an option for enterprises seeking greater control over their AI stacks.

Palantir Chief Executive Alex Karp used his quarterly shareholder letter to warn about AI agents deployed in enterprises without Palantir’s technology. He said agents without proper context, grounding and safeguards create potentially unbounded liability.

“the LLM agent is a virtually unlimited liability, possessing a nearly unbounded destructive potential and a blast radius that tends towards infinity,” Karp wrote. “There is simply no means of effectively and reliably connecting the language models to large enterprises without the operating model and capabilities we have pioneered.” Increasingly, Karp added, business leaders in the U.S. are waking up to the need for context, grounding and safeguards when enterprises deploy AI agents.

Karp also said businesses were increasingly being distinguished by their control over internal AI stacks, a condition known as “Sovereign AI.”

“The Sovereign AI movement is imposing the most dramatic dichotomy between haves and have nots that I have witnessed in 30 years of business,” Karp said. “Every institution—government and commercial—is rapidly being forced, voluntarily or not, into one of these two categories.” Palantir has consistently argued that businesses are better off using Palantir to build proprietary software with open-weight AI models than relying on a third-party AI provider such as Anthropic or OpenAI.

StoneX analysts projected that top AI hyperscalers and neoclouds would post much wider free-cash losses next year. Their review covered six leading hyperscalers along with CoreWeave and Nebius, and estimated combined group losses of more than $309 billion in 2027, compared with more than $112 billion this year.

The analysts said the inability to fund the free-cash-flow deficit over the next few years represented the single greatest concern on Wall Street at the moment. Until investors became more confident about funding the deficit, or more willing to fund it, chip companies with heavy exposure to the data center market could struggle to re-rate higher.

Across the market notes, companies and analysts tracked related financial questions from different directions: whether Tencent could earn returns on AI spending, whether Apple could absorb memory and Siri AI costs, how Xiaomi’s smartphone revenue compared with pressure in its new initiatives, and whether six hyperscalers, CoreWeave and Nebius could secure enough capital to bridge their projected combined losses.