Tech Express | July 17, 2026 Global Chip Selloff: AI Valuation Fears Trigger Semiconductor Rout, Asian Markets Hammered

---

📰 Global Chip Stocks Bloodbath: SK Hynix Tumbles 11%, Nikkei Drops Over 5%, KOSPI Enters Bear Market

On July 16-17, global semiconductor stocks faced a brutal selloff as fears over AI hardware overvaluation came to a head. Investors grew increasingly skeptical that tech giants' massive capital expenditures could translate into near-term profits. SK Hynix shares plunged 11.53% in Seoul—the steepest single-day drop this year—while South Korea's KOSPI index crashed over 6%, officially entering bear market territory.

Japan's market was similarly devastated, with the Nikkei 225 tumbling more than 5% intraday on July 17, shedding over 2,000 points at one stage. Taiwan's weighted index found some support from TSMC's stellar Q2 earnings, but still couldn't withstand relentless foreign sell pressure. Notably, South Korea's market was closed for a national holiday on July 17, narrowly avoiding a second wave of carnage.

The immediate trigger was news that Chinese memory chip maker CXMT is preparing an $8.6 billion IPO, stoking fears of a structural supply glut in the global memory chip market. Following SK Hynix's lead, SanDisk dropped over 12%, Micron fell more than 5%, and Western Digital slid over 9%, wiping out hundreds of billions in market value across the sector.

— CNBC

---

📰 Apple Intelligence Cleared for China Launch, Partners with Alibaba's Qwen AI

On July 15, China's Cyberspace Administration published its latest batch of generative AI service registrations, confirming that Apple Technology Development (Shanghai) Co.'s "Apple Intelligence" has completed regulatory filing for use on domestic iPhones. From the initial WWDC 2024 announcement to full approval, Chinese users have waited two full years.

Concurrently, Apple has struck a deal with Alibaba to integrate the Qwen large language model into Apple Intelligence across iOS, iPadOS, macOS, and visionOS. Alibaba shares rose on the news. Six other AI services—mostly from Chinese smartphone manufacturers—were also approved in the same batch.

This marks the clearing of the biggest regulatory hurdle for Apple's AI strategy in the world's second-largest smartphone market. Unlike its US deployment integrating OpenAI's ChatGPT, the China version of Apple Intelligence leans on a local partner to comply with data sovereignty and regulatory requirements.

— Reuters

---

📰 SpaceX Scrub: Starship Flight 13 Aborted at Last Second Before Liftoff

On July 16, SpaceX was set to launch the 13th integrated test flight of its Starship megarocket from Starbase, Texas, when the automated system triggered an abort right before liftoff. This was to be the second flight of the Starship V3 configuration, originally planned to deploy the first functional Starlink simulator payload and gather critical data for NASA's Artemis lunar lander program.

Elon Musk explained on X: "To be confident of a good flight, 2 Raptors will be removed & replaced. Most probable launch timing is early next week." SpaceX engineering teams are currently assessing the specific technical cause, and a new launch date has not yet been formally announced.

The scrub carries significant implications for NASA, as Starship serves as the human landing system for the Artemis program and its progress directly affects the US goal of returning astronauts to the Moon by 2030. CNN reports that NASA is facing increasingly tight deadline pressure.

— CNN

---

📰 Apple Sues OpenAI Over Trade Secret Theft by Former Employees

On July 10, Apple filed a 41-page complaint in the US District Court for the Northern District of California, formally suing OpenAI and its Chief Hardware Officer Tang Tan. The lawsuit alleges that multiple former Apple engineers, upon leaving to join OpenAI, stole vast amounts of sensitive technical documents and trade secrets.

The complaint's details are striking: Apple accuses former employees not only of stealing software-level design files but also of attempting to gain unauthorized physical access to Apple hardware. Apple claims these trade secrets involve core technologies for AI devices critical to the iPhone maker's future product roadmap.

OpenAI subsequently issued a formal statement denying Apple's allegations. Wedbush analysts noted that the lawsuit underscores the strategic importance of AI hardware to Apple's future—the company is deeply integrating AI into its next-generation products, including the rumored foldable iPhone 18 Ultra.

— 9to5Mac

---

📰 Microsoft Launches Frontier Company: $2.5 Billion and 6,000 Engineers to Fix AI's ROI Problem

Microsoft Commercial Business CEO Judson Althoff announced on July 2 the creation of Microsoft Frontier Company, a new subsidiary backed by $2.5 billion and 6,000 AI engineers embedded directly inside enterprise customer organizations. The unit's core mission: fixing the "ROI problem" that has plagued enterprise AI adoption.

Frontier Company employs a "forward-deployed engineering" model where engineers co-design, deploy, and continuously iterate AI systems on-site alongside customer teams, with contracts tied to measurable business outcomes rather than traditional time-and-materials billing. The company operates on a dual-platform architecture—an "intelligence platform" that compounds customer proprietary data and decision-making, and a "trust platform" handling governance, security, and FinOps-based ROI tracking.

Microsoft makes an explicit intellectual property protection commitment: customer data, proprietary processes, and competitive intelligence will not be used for shared model training—directly addressing CIOs' and CISOs' top concern. Early customers include LSEG (London Stock Exchange Group), Land O'Lakes, Unilever, and Novo Nordisk, with Accenture, EY, Capgemini, KPMG, and PwC signed on as global SI deployment partners.

— MarketScale

---

📰 WAIC 2026 Opens in Shanghai: 108 Chips, 261 LLMs, Over 300 Robots on Display

From July 17-20, the 2026 World Artificial Intelligence Conference and High-Level Meeting on Global AI Governance (WAIC 2026) opens at the Shanghai Expo Center. Chinese President Xi Jinping attended the opening ceremony and delivered a keynote address, signaling China's ambition to play a leadership role in global AI governance.

This year's WAIC is packed with cutting-edge exhibits: 108 chips, 261 large language models, 208 embodied AI terminals, and over 300 live robots on display. The conference program features more than 140 industry forums covering AI safety, LLM applications, autonomous driving, and embodied intelligence. International organizations including the PyTorch Foundation are also participating.

The concurrent High-Level Meeting on Global AI Governance addresses cross-border AI regulatory frameworks, data flows, and safety alignment. Against the backdrop of the global chip selloff and growing AI investment return skepticism, WAIC 2026 serves as a critical window into China's AI industrial strategy and global cooperation trajectory.

— 36Kr

---

📰 Stripe and Advent Make $53 Billion Unsolicited Bid for PayPal, Board Says It's Too Low

According to Reuters sources, payments giant Stripe and private equity firm Advent International have jointly submitted an unsolicited $53 billion takeover offer for PayPal at $60.50 per share. PayPal shares surged as much as 17% on the news.

However, PayPal's board believes the offer undervalues the business, and negotiations are ongoing. If completed, this would be the largest acquisition in global fintech history. The bid is seen as Stripe's strategic move to pursue scale advantages and expand its global footprint amid broader payments industry consolidation.

Analysts note that PayPal commands a massive user base and merchant network, with footholds in emerging areas like stablecoins. Stripe and Advent may need to sweeten the offer or attach more favorable terms to win over PayPal shareholders.

— Reuters via Private Equity Wire

---

📰 Meta Rolls Out Teen AI Safety Alerts: Self-Harm Discussions Will Notify Parents

On July 16, Meta announced a new safety feature for its Meta AI chatbot: when the system detects that a teen user's conversation involves self-harm or suicidal ideation, it will automatically send an alert to parents or guardians.

This builds on Meta's earlier initiative to show guardians the topics of their teens' AI conversations, launched a few months ago. The new feature adds a human review layer—trained safety teams verify AI-detected alerts to ensure accuracy and appropriateness. The feature is rolling out in phases across regions where Meta AI is available.

Online safety organizations including ConnectSafely have expressed support for Meta's move, viewing it as an important step in translating AI safety responsibility into concrete product features. However, privacy advocates have raised concerns that such monitoring mechanisms could impact teens' freedom of expression.

— Engadget

---

📰 Mozilla Investigation Reveals Period Tracker Privacy Black Hole: Apps Sharing Sensitive Health Data

BBC Future, in collaboration with the Mozilla Foundation, has published a groundbreaking investigation revealing disturbing data-sharing practices among popular period-tracking apps. The research found that multiple widely used period trackers share sensitive health data with third-party analytics firms and advertisers without users' explicit informed consent.

Among the apps named is Stardust, which was found to send user data to analytics platform RudderStack. While a Stardust spokesperson stated that RudderStack is "contractually prohibited from selling or using it for its own purposes," both are US-based companies that could still receive data demands from government or law enforcement agencies. In an increasingly sensitive political environment surrounding women's health data, the findings have sparked widespread concern.

The research also highlights apps with strong privacy practices, helping users make more informed choices. The report calls on regulators to strengthen scrutiny of health app data handling and advises users to carefully review privacy policies when choosing period-tracking tools.

— BBC

---

📰 Anthropic Invests $10 Million in Canadian AI Research, Claude's "Hidden Canadian Roots" Revealed

BetaKit reports that AI giant Anthropic—now valued at over CAD $1 trillion—announced this week a $10 million investment in Canadian AI research. The funding will be distributed as research grants and Claude model credits to support academic institutions and independent researchers across Canada.

Vass Bednar, Managing Director of Canadian think tank Canadian Shield Institute, quipped: "That is, if you count Claude credits as investment." The remark reflects some industry skepticism about AI companies' "credits-in-lieu-of-cash" sponsorship model, but also underscores Canada's significant position in the global AI research ecosystem.

BetaKit's accompanying feature article, "Which LLMs Are Secretly Canadian?", explores the deep contributions of Canadian talent and technology to global LLM development—from pioneering research at the University of Toronto and Université de Montréal to the profound influence of "AI godfathers" like Geoffrey Hinton, Canada has long been an AI talent powerhouse.

— BetaKit