As US and Chinese officials prepare for high-stakes AI talks next month, and ransomware attacks continue to halt production at major consumer brands, the technology world is navigating a tense mix of diplomatic breakthroughs, competitive pressures, and real-world disruptions.
The battle for control of the AI economy is no longer playing out inside research labs alone. It is moving into diplomatic talks, export restrictions, chip factories, copyright courts, data-center financing, and the payment systems that connect billions of people. Today’s biggest tech stories reveal an industry being reshaped from the ground up, as governments tighten their grip, investors pour billions into infrastructure, and companies race to secure the chips, energy, data, and legal rights needed to compete.
It’s Tuesday, July 21, 2026, and here are the top tech news stories shaping the global tech and startup landscape today — from frontier AI models and regulatory shifts to cybersecurity threats hitting supply chains and the rise of synthetic content in advertising.
Technology News Today
US and China plan first official AI talks under Trump administration
The United States and China are preparing to hold formal talks on artificial intelligence in September, opening a new diplomatic channel focused on the growing security and economic risks surrounding frontier AI models. The discussions are expected to take place before Chinese President Xi Jinping’s planned September 24 visit to the United States, although the location, agenda, and full list of participants have not been finalized.
The talks are expected to cover military uses of AI, cyberattacks against critical infrastructure, access to advanced models, and the release of increasingly capable open-weight systems. Washington has restricted exports of advanced AI chips to China and is considering further controls on the use of Chinese models by American companies. Beijing, meanwhile, is weighing its own restrictions on overseas access to advanced Chinese models and sensitive training data.
The meeting would be the first official U.S.-China AI dialogue under President Donald Trump. It is unlikely to produce a broad agreement, but establishing shared definitions for frontier models, proliferation risks, and model-release standards could become an important first step. The negotiations also show that AI is moving beyond trade and industrial policy into the same diplomatic category as nuclear technology, cybersecurity, and military arms control.
Why It Matters: The world’s two largest AI powers are beginning direct negotiations over technologies that could reshape national security, cyberwarfare, and the global economy.
Source: Reuters.
Tesla’s cash burn puts its expensive AI and robotics bets under scrutiny
Tesla’s spending on artificial intelligence, autonomous driving, robotics, and computing infrastructure is expected to face renewed scrutiny as investors assess whether the company’s emerging technology businesses can justify their rising costs. Tesla has positioned robotaxis, the Optimus humanoid robot, and its in-house AI systems as central to its long-term valuation, even as its core electric-vehicle business faces stronger competition.
Building those businesses requires large capital commitments. Tesla must purchase chips, expand data centers, train increasingly large driving models, manufacture experimental hardware, and deploy fleets of vehicles for testing and commercial service. These investments arrive while automotive pricing pressure and manufacturing expenses continue to affect cash generation.
Supporters argue that Tesla’s vehicle fleet gives it access to driving data that rivals cannot easily replicate. Critics question how quickly that advantage can translate into safe, widely approved autonomous services. Robotaxi deployment is also shaped by regulators, insurance requirements, geographic limitations, and the cost of maintaining vehicles.
Tesla’s financial results will therefore be judged not just on vehicle deliveries but on whether its AI programs are producing measurable commercial progress rather than distant promises.
Why It Matters: Tesla’s valuation increasingly depends on AI and robotics businesses that require heavy spending before they can generate meaningful revenue.
Source: CNBC.
AI-generated fake influencers flood social media with misleading wellness supplement ads
The New York Times investigated hundreds of hyper-realistic AI-generated avatars posing as doctors, healers, and wellness influencers on platforms like Facebook, Instagram, and TikTok. These synthetic figures promote health supplements with before-and-after transformations and bold claims, often targeting older women with health issues. Examples include repeated imagery of American flags and suburban settings, fake diplomas, and assertions that herbal products outperform medications for conditions like kidney disease.
One case involved Rosabella moringa supplements linked to a Salmonella recall by the FDA and CDC, with consumer complaints rising. Chinese agencies are producing up to 1,200 such AI videos daily at low cost for U.S. markets via TikTok Shop. Platforms label or remove some content, but enforcement lags due to the speed of AI tools and anonymous LLC operators. No federal U.S. law currently restricts synthetic people in advertising. This development highlights how accessible AI video generation is accelerating disinformation in consumer marketing while outpacing moderation capabilities. It raises urgent questions about trust in digital platforms, consumer protection, and the global supply chain of synthetic content originating from regions with fewer regulations.
Why It Matters: AI tools are enabling scalable misleading advertising that exploits vulnerable consumers and erodes trust in digital ecosystems, underscoring the need for stronger detection, labeling, and enforcement mechanisms across platforms and regulators.
Source: New York Times.
China considers export controls on advanced AI models, chips, and training data
China is considering tighter restrictions on exports of advanced AI models, semiconductor technologies, and critical training data as Beijing seeks to protect technologies it increasingly views as strategic national assets. Chinese regulators have reportedly consulted major domestic technology companies, including Alibaba, ByteDance, and Zhipu AI, about possible rules governing foreign access to powerful models and the data used to develop them.
The proposed controls could be added to China’s catalog of restricted export technologies. Officials are also examining whether overseas acquisitions, joint ventures, or manufacturing arrangements could allow foreign companies to gain control of strategically important Chinese AI or semiconductor intellectual property. The discussions remain preliminary, and no final policy has been announced.
The move would represent a major reversal in the technology rivalry between China and the West. For years, Washington has used export controls to restrict China’s access to Nvidia AI chips, advanced semiconductor equipment, and chipmaking software. Beijing is now signaling that it may use similar tools to prevent Chinese AI systems and designs from strengthening foreign competitors. The potential rules could affect cloud providers, model developers, chip manufacturers, and companies building products around Chinese open-weight AI.
Why It Matters: China is preparing to treat advanced AI models and semiconductor designs as controlled strategic exports, adding another layer to the global technology divide.
Source: Financial Times.
BlackRock and MGX commit another $5 billion to Aligned Data Centers
A consortium backed by BlackRock and Abu Dhabi investment firm MGX has committed an additional $5 billion in growth capital to Aligned Data Centers after completing its acquisition of the company. The deal valued Aligned at roughly $40 billion and places one of the world’s largest data-center operators under the control of investors building a broad global AI infrastructure portfolio.
Aligned operates or is developing 51 data-center campuses with more than 6.4 gigawatts of current and planned capacity. Its facilities serve hyperscalers, cloud providers, and companies deploying large AI computing clusters. The new capital is expected to accelerate construction as demand for electricity, cooling systems, fiber connectivity, and high-density server space continues to grow.
The consortium has discussed deploying as much as $30 billion in equity, with total investment capacity potentially reaching $100 billion when debt financing is included. The scale shows how data centers are becoming an institutional asset class comparable to energy, transportation, and telecommunications infrastructure. Investors are no longer betting solely on individual AI model developers. They are increasingly financing the physical facilities that nearly every model company, cloud platform, and enterprise AI service will need.
Why It Matters: The additional investment shows that global capital is treating AI data centers as foundational infrastructure rather than a temporary technology spending cycle.
Source: BlackRock and Global Infrastructure Partners.
Anthropic’s $1.5 billion AI copyright settlement receives final court approval
A federal judge in San Francisco has signed off on Anthropic’s $1.5 billion settlement with authors who accused the AI company of obtaining pirated books and storing them in a central digital library. The case centered on millions of books collected as Anthropic developed Claude and its underlying language models.
The court previously drew an important distinction between using lawfully obtained books to train AI models and acquiring unauthorized copies from piracy websites. While the judge found that some forms of AI training could qualify as fair use, Anthropic still faced potentially enormous liability over how it obtained and stored the copyrighted material. The settlement removed the threat of a trial in which statutory damages could have reached far beyond $1.5 billion.
More than 91% of eligible authors and publishers have reportedly submitted claims. The final approval also includes more than $101 million in attorneys’ fees. The ruling does not resolve every AI copyright dispute, and separate lawsuits against Anthropic, OpenAI, Meta, and other model developers remain active. However, it gives publishers, authors, and AI companies a clearer view of where courts may draw the line between transformative model training and unauthorized acquisition of copyrighted datasets.
Why It Matters: AI companies may have broad arguments for training on copyrighted works, but those protections do not erase liability for obtaining the material through piracy.
Source: Fundpluse via Reuters, U.S. District Court for the Northern District of California.
TSMC reportedly plans chip price increases of up to 10% in 2027
Taiwan Semiconductor Manufacturing Company is preparing to raise chipmaking prices by as much as 10% beginning in 2027, according to a report from Nikkei Asia. The proposed increases would help the world’s largest contract chipmaker absorb higher material costs, more expensive manufacturing equipment, and the cost of building semiconductor plants outside Taiwan.
The reported increases would affect both advanced and mature manufacturing processes. That matters because TSMC produces chips for a broad range of customers, including Nvidia, Apple, AMD, Qualcomm, and many smaller semiconductor companies that rely on external manufacturing. Even modest price changes can move through the technology supply chain and eventually affect the cost of AI accelerators, smartphones, servers, networking equipment, and consumer electronics.
TSMC is investing heavily in new facilities in the United States, Japan, Germany, and Taiwan while simultaneously developing increasingly expensive manufacturing processes. Advanced lithography equipment can cost hundreds of millions of dollars per machine, and new fabrication plants can require tens of billions of dollars. Customers may accept higher prices because relatively few alternatives can match TSMC’s manufacturing scale, yields, and ability to produce the most advanced processors.
Why It Matters: Higher TSMC prices could raise costs across the global technology industry, especially for companies building advanced AI chips and high-performance computing systems.
Source: Nikkei Asia.
Ant International raises $1.2 billion to expand global fintech and AI commerce
Ant International has raised approximately $1.2 billion in a new equity financing round as the Singapore-based fintech company expands its payment and enterprise financial services outside China. Investors included Ant Group, Alibaba, and other participants, although the companies did not disclose the size of each investment.
The company operates payment, account-management, merchant-finance, and cross-border transaction services across Asia, Europe, Latin America, and the Middle East. Ant International says its network connects approximately 150 million merchants with around two billion user accounts through partnerships with banks, payment systems, digital wallets, and technology platforms.
The round also gives Ant International more capital to develop services for what the company calls agentic commerce, where AI assistants may eventually search for products, compare offers, authorize transactions, and manage payments on behalf of users or businesses. That future will require trusted systems for identity, fraud prevention, settlement, and cross-border compliance. Ant International is positioning itself as part of that financial infrastructure.
The financing also marks a new chapter for businesses created through Ant Group’s restructuring following the regulatory crackdown that halted its planned 2020 IPO.
Why It Matters: Ant International is building a global payment layer for conventional digital commerce and a future in which AI agents increasingly initiate transactions.
Source: The Next Web.
Chinese AI infrastructure supplier Zhongji Innolight targets $8 billion Hong Kong IPO
Zhongji Innolight, a Chinese manufacturer of optical transceivers used in AI data centers, is preparing to raise at least $8 billion through a Hong Kong listing. At that size, the offering would rank among Asia’s largest technology IPOs and become Hong Kong’s biggest listing since Alibaba’s secondary offering in 2019.
Optical transceivers transfer data between servers, switches, and computing clusters. Their importance has increased as AI training systems have grown from hundreds to tens of thousands of interconnected chips. A cluster’s performance depends not just on the processors but on how quickly data can move between them without creating network bottlenecks.
Zhongji’s revenue reportedly rose 192% from a year earlier during the first quarter of 2026, while profit climbed 274%. The United States accounted for more than 60% of its quarterly revenue, highlighting the complicated commercial ties that remain between Chinese component suppliers and American AI infrastructure operators despite widening geopolitical restrictions.
The company plans to use the proceeds for research, acquisitions, supply-chain expansion, and international growth. The offering could also test whether investors remain willing to back Chinese technology companies with significant exposure to the U.S. AI market.
Why It Matters: The planned IPO shows how the AI infrastructure boom is creating multibillion-dollar opportunities for networking suppliers, not just chipmakers such as Nvidia.
Source: South China Morning Post.
France moves toward banning social media for children under 15
French lawmakers are voting on legislation that would prohibit children under the age of 15 from opening accounts on major social-media platforms. The proposed restrictions would apply to services including TikTok, Instagram, Facebook, Snapchat, and YouTube, while educational websites and online encyclopedias would remain available.
Under the legislation, platforms would need to introduce age-verification systems approved by France’s privacy regulator. New underage accounts could be prohibited beginning September 1, while companies would receive additional time to identify and close existing accounts held by children below the minimum age.
The French proposal follows Australia’s introduction of restrictions for users under 16 and comes as governments across Europe examine how algorithmic feeds, targeted advertising, and engagement-based product design affect young users. Technology companies generally argue that blanket bans may be difficult to enforce and could push teenagers toward less regulated services. Supporters counter that existing parental controls and voluntary platform safeguards have failed to address addiction, exposure to harmful content, and the collection of children’s data.
European Union regulators will also assess whether the French law is compatible with existing regional digital-platform rules.
Why It Matters: France could push age verification from a limited safety feature into a legal requirement for some of the world’s largest social-media platforms.
Source: France 24.
Chinese AI models deliver one-two punch to U.S. dominance claims
Recent releases from Chinese labs, including Moonshot AI’s Kimi K3 and Alibaba’s Qwen series, have demonstrated competitive performance against leading U.S. frontier models at significantly lower costs. These open-weight systems are gaining traction among U.S. companies for inference and fine-tuning, narrowing the perceived capability gap. Observers note that China’s rapid progress in open models challenges assumptions of sustained American leadership in advanced AI.
The developments underscore how cost advantages and open approaches are reshaping global AI accessibility. U.S. firms and policymakers are increasingly grappling with competitive pressures from more affordable, capable alternatives.
Why It Matters: Chinese AI advancements are eroding U.S. technological edge in frontier models, forcing Big Tech and startups to reassess sourcing strategies, costs, and national security considerations in AI infrastructure.
Source: The Verge.
US government seeks faster approval path for air taxis and supersonic aircraft
The United States is moving to shorten the regulatory path for electric air taxis and new supersonic passenger aircraft as policymakers try to prevent emerging aviation companies from shifting development and testing overseas. The effort focuses on accelerating certification, infrastructure planning, and commercial deployment while preserving the Federal Aviation Administration’s safety requirements.
Electric vertical takeoff and landing aircraft, commonly known as eVTOLs, are being developed by startups and established aerospace companies for short urban and regional flights. Several manufacturers have already conducted extensive test programs, but certification timelines, pilot requirements, noise rules, battery standards, and the construction of landing facilities remain significant barriers.
The government is also revisiting rules affecting civil supersonic aviation. New aircraft developers argue that better engines, lighter materials, and improved aerodynamic designs could make faster-than-sound travel quieter and more economical than earlier generations of supersonic jets. Regulators must still address sonic booms, airport noise, emissions, and international operating standards.
Faster approvals could help U.S. aviation startups reach commercial service before overseas competitors. However, any effort to reduce delays will be closely watched after recent scrutiny of aviation manufacturing and regulatory oversight.
Why It Matters: Certification speed may determine whether the next generation of air-taxi and supersonic companies scales in the United States or moves investment to more accommodating markets.
Source: U.S. Department of Transportation.
New York receives $400 million ASML chipmaking system for advanced semiconductor research
The first components of an advanced ASML High NA extreme ultraviolet lithography system have arrived at the Albany NanoTech Complex in New York. The approximately $400 million machine will support research into future semiconductor manufacturing processes as chipmakers attempt to build smaller, faster, and more energy-efficient processors.
High NA EUV lithography represents the next major generation of the technology used to print tiny circuit patterns onto silicon wafers. The system uses a higher numerical aperture than current EUV equipment, allowing manufacturers to produce finer features with fewer processing steps. Intel has already installed early High NA equipment, while other manufacturers are evaluating when the technology becomes economically necessary.
The Albany facility is operated by NY Creates in partnership with companies including IBM, Micron, and Tokyo Electron. It is one of North America’s most important collaborative semiconductor research centers and is often compared with Imec in Belgium. Additional components will arrive over the coming weeks, with the machine expected to become operational by the end of the year.
The installation gives U.S. researchers access to equipment that will help shape manufacturing technologies beyond today’s leading production nodes.
Why It Matters: Access to High NA EUV technology strengthens America’s ability to conduct advanced chip research without relying entirely on overseas semiconductor laboratories.
Source: NY Creates.
Brazil’s Pix payment system becomes a new front in US technology tensions
Brazil and the United States are clashing over Pix, the instant-payment system created by Brazil’s central bank. Pix has become deeply embedded in the country’s economy, allowing consumers and businesses to send money immediately using mobile apps, QR codes, phone numbers, and other simple identifiers.
Its growth has reduced reliance on cash, card networks, and some payment services operated by American companies. That success has drawn attention in Washington, where officials and private-sector groups are examining whether government-supported payment systems create unfair competition for U.S. financial and technology companies.
Brazilian officials argue that Pix is public financial infrastructure that has lowered transaction costs, increased competition, and brought more people into the formal financial system. The dispute reflects a wider struggle over who will control the next generation of payments. Countries in Asia, Latin America, Africa, and Europe are developing domestic instant-payment networks and connecting them across borders.
For finFundpluse, these systems create opportunities to build merchant tools, lending services, fraud detection, and financial applications. For established card networks and payment platforms, they threaten transaction fees and the control that comes with owning the payment rails.
Why It Matters: Pix shows how government-built digital infrastructure can challenge dominant private payment networks and become a source of international technology friction.
Source: Rest of World.
AI could lift Sub-Saharan Africa’s economy by 4%, but infrastructure remains the barrier
Artificial intelligence could increase economic output across Sub-Saharan Africa by roughly 4% if countries improve access to reliable electricity, internet connectivity, computing infrastructure, and digital skills, according to new analysis highlighted by the International Monetary Fund.
The region has a large young population and growing mobile adoption, creating opportunities for AI applications in agriculture, healthcare, education, financial services, and government administration. Local-language models could help users access services in communities where major global AI platforms have limited linguistic coverage. AI tools may also help small companies automate accounting, customer service, logistics, and market research.
The potential gains are constrained by persistent infrastructure gaps. Many communities lack dependable electricity or affordable broadband, while governments and startups have limited access to the computing resources required to train and deploy AI systems. Weak data protection, shortages of technical talent, and dependence on foreign cloud providers create additional challenges.
The analysis suggests that AI policy in Africa cannot focus solely on model adoption. Investment in energy grids, connectivity, education, and locally relevant datasets will determine whether the technology broadens economic participation or concentrates opportunity in a small number of connected urban centers.
Why It Matters: Africa’s AI opportunity depends less on access to chatbots than on electricity, affordable internet, skills, and infrastructure that allow the technology to reach businesses and communities.
Source: International Monetary Fund.
Mobileye will supply Stellantis with cloud-based driver-assistance technology
Mobileye has reached an agreement to provide Stellantis with cloud-connected driver-assistance technology, extending the chip and autonomous-driving company’s relationship with one of the world’s largest automakers. The system is expected to combine in-vehicle sensors and processors with software services that can receive updates, analyze driving information, and improve assistance features over time.
Stellantis owns brands including Jeep, Ram, Chrysler, Peugeot, Fiat, Citroën, and Opel. Deploying a shared driver-assistance platform across multiple brands could help the automaker reduce development costs and introduce features more consistently across different vehicle models and geographic markets.
For Mobileye, the deal supports its strategy of selling a unified technology stack that includes computer vision, mapping, driving policy, specialized processors, and mathematical safety systems. The company is competing with Nvidia, Qualcomm, Tesla, and automakers developing more of their driving software internally.
Cloud connectivity is becoming a central part of the automotive technology business. Vehicles increasingly receive software updates after purchase, while manufacturers collect data to diagnose problems and refine automated-driving systems. That shift also raises questions about cybersecurity, data ownership, subscription pricing, and how long automakers will support connected features.
Why It Matters: The agreement shows that automakers are treating driver assistance as an updatable software platform rather than a fixed feature installed when a vehicle leaves the factory.
Source: Mobileye.
UK chip supplier IQE raises forecast as AI data-center demand boosts sales
British semiconductor wafer manufacturer IQE has raised its full-year sales growth forecast after stronger-than-expected first-half trading, supported by demand from AI infrastructure and data-center customers. IQE produces specialized semiconductor wafers used in communications equipment, sensors, radio-frequency components, and advanced computing systems.
The company operates further upstream than better-known AI chipmakers. Its materials become part of components used in optical connections, wireless systems, and other hardware that allows processors, servers, and data centers to communicate. That position gives IQE exposure to the infrastructure buildout without competing directly in the market for graphics processors or AI accelerators.
Its improved outlook adds to evidence that AI investment is moving through multiple layers of the semiconductor supply chain. Nvidia and major cloud providers may capture much of the attention, but increased spending also benefits wafer producers, memory manufacturers, optical-networking companies, cooling suppliers, equipment makers, and power-management chip companies.
For the United Kingdom, IQE is also strategically important as one of the country’s few publicly traded semiconductor-materials companies. Its performance will influence the broader debate over whether Britain can preserve meaningful domestic capabilities in advanced chip technologies.
Why It Matters: AI data-center spending is lifting demand across the semiconductor supply chain, including specialized materials suppliers that sit far behind the industry’s most visible chipmakers.
Source: IQE.
That’s your quick tech briefing for today. Follow us on X @TheFundpluse for more real-time updates.
