Tech News Today, October 22, 2026: Barclays, DeepSeek, IBM, Google, Meta, Nokia, Palantir, Tesla & More

Daily Global Technology Digest | Thursday, October 22, 2026

It’s Thursday, October 22, 2026, and the tech industry’s contradictions are becoming harder to ignore. AI investment continues to soar, with DeepSeek reportedly nearing a $15 billion funding round and Meta reporting $64.6 billion in quarterly revenue. Yet beneath those numbers, new pressures are emerging: electricity constraints are disrupting data-center expansion, cybersecurity threats are escalating, and a $25 billion judgment against Meta shows that regulatory risks can carry enormous financial consequences.

The race for AI leadership is also taking different paths. Chinese AI startup DeepSeek is attracting backing from some of the country’s largest technology and industrial companies, while American research startup Flapping Airplanes has raised $1 billion to explore whether advanced models can be developed using substantially less computing power. Meanwhile, Swedish legal AI startup Legora has secured $550 million at an $8.8 billion valuation, demonstrating continued investor demand for specialized AI applications with established enterprise customers.

The financial picture is becoming increasingly complicated. Meta’s advertising business continues to generate substantial revenue, but its enormous AI infrastructure spending is pressuring cash flow. In New Mexico, a judge has ordered the company to pay $25 billion in consumer protection penalties, a ruling Meta plans to appeal. Elsewhere, Nokia’s latest earnings show how AI data-center demand is creating new opportunities for networking equipment suppliers even as electricity shortages and construction delays threaten other infrastructure investments.

Cybersecurity remains another major concern. Researchers report widespread attempts to exploit two critical PHP vulnerabilities after Wednesday’s emergency patches. Google is strengthening Chrome against malicious websites that trick users into executing dangerous commands, while Ukrainian drone attacks have damaged Russian data centers supporting digital services and AI research. The incidents highlight how both software weaknesses and physical infrastructure attacks can disrupt increasingly interconnected technology systems.

The consequences extend beyond computing infrastructure. British bank Barclays is cutting approximately 5,000 jobs as it pursues greater efficiency through technology and AI, while Tesla is balancing investments in autonomous vehicles and robotics against declining quarterly profits. IBM is reporting stronger software revenue, OPPO is introducing its latest flagship smartphones, and Northwestern researchers have demonstrated more precise control over a single particle of antimatter.

The bigger picture is becoming clearer: AI is creating extraordinary opportunities, but building more capable models is only part of the challenge. The next phase will be shaped by access to electricity, sustainable financing, secure infrastructure, regulatory accountability, and the ability to turn technological advances into measurable economic value.

From DeepSeek’s massive fundraising and Meta’s $25 billion legal battle to emerging cyber threats, AI infrastructure constraints, and breakthroughs in fundamental physics, here are the top tech news stories shaping the day.

New Mexico Judge Orders Meta to Pay $25 Billion in Landmark Consumer Protection Case

A New Mexico judge ordered Meta to pay $25 billion in penalties Wednesday after a jury found that Facebook repeatedly violated the state’s consumer protection law. First Judicial District Judge Francis Mathew issued the judgment following a September jury verdict involving nearly 44 million violations of New Mexico’s Unfair Practices Act. The case examined Facebook’s representations about personal information, content moderation, community standards, and its response to the Cambridge Analytica scandal. State Attorney General Raúl Torrez described the judgment as a major consumer victory. The court concluded that Meta’s conduct warranted substantial penalties, although the amount awarded was considerably below the theoretical maximum available under the statute. Meta disputes the ruling and says it expects to prevail on appeal. The $25 billion figure is therefore a court-ordered penalty, not money New Mexico has already collected.

The judgment could influence how technology companies assess legal exposure arising from statements about privacy and platform safety. Consumer protection authorities increasingly scrutinize whether companies accurately describe their data practices and whether advertised safeguards operate as users reasonably expect. The penalty’s scale also shows how relatively small jurisdictions can pursue significant claims against global technology companies when a large number of alleged violations accumulate. However, the legal outcome remains subject to appellate review, and the eventual financial consequences could differ substantially from the current judgment. This case is separate from the multistate youth-related litigation reflected in Meta’s latest earnings. For startups and established platforms, the broader lesson concerns the importance of ensuring that privacy policies, user controls, and public representations match actual product behavior.

Why It Matters: The $25 billion judgment raises the financial stakes for technology companies facing claims that their privacy and consumer protection statements are misleading.

Source: Source New Mexico.

DeepSeek Nears $15 Billion Funding Round as China’s AI Giants Challenge OpenAI and Anthropic

Chinese artificial intelligence startup DeepSeek is nearing completion of a financing round worth approximately 100 billion yuan, or $15 billion, as investors bet on its ability to compete with leading American AI developers. Reuters reported Thursday that the round includes backing from major Chinese companies, including battery manufacturer CATL, Tencent, Meituan, Geely, GigaDevice, and several investment firms. DeepSeek is reportedly valued at approximately 500 billion yuan before the new investment, equivalent to roughly $75 billion. The financing would double the company’s original target of 50 billion yuan. However, people familiar with the transaction told Reuters that the final investor lineup and financial terms could still change before the round closes. The funding should therefore not be treated as a fully completed transaction.

The investment reflects China’s efforts to build a competitive domestic AI ecosystem while facing American restrictions on advanced semiconductor exports. DeepSeek has attracted international attention with models that offer an alternative to proprietary systems from OpenAI, Anthropic, and Google. Its emphasis on downloadable models has also helped developers experiment with AI without relying exclusively on American cloud providers. The reported participation of Chinese industrial and technology companies could give DeepSeek strategic relationships beyond financial capital, including opportunities in computing infrastructure and commercial deployment. Reuters also reported that the financing could support preparations for a domestic initial public offering. For competitors, the scale of the proposed investment illustrates how quickly the market for leading AI developers is becoming concentrated around companies with access to enormous capital reserves.

Why It Matters: DeepSeek’s reported financing could strengthen China’s domestic AI industry while intensifying competition with American frontier-model developers.

Source: TechStartups via Reuters.

Tesla Q3 Profit Falls 44% to $787 Million Despite Positive Cash Flow as Musk Pushes AI and Robotaxi Strategy

Tesla reported a 44% decline in third-quarter net income Wednesday, generating $787 million in profit despite stronger-than-expected cash flow. The electric vehicle manufacturer earned approximately 22 cents per share under standard accounting measures, while adjusted earnings of 48 cents per share exceeded analysts’ expectations. Tesla also reported roughly $1.5 billion in free cash flow, an important development because investors had expected spending on AI, robotics, and autonomous driving to consume more cash than the company generated. Chief Executive Elon Musk continues positioning Tesla as a business extending beyond electric vehicles into autonomous transportation, robotics, and AI infrastructure. However, the latest results show that those ambitions must be financed alongside an automotive operation facing pricing pressure and intense competition.

Tesla’s financial performance matters because the company is trying to develop several capital-intensive technologies at once. Building humanoid robots, training autonomous driving systems, and expanding transportation services all require significant engineering investment and computing resources. The company must also maintain vehicle manufacturing capacity and support its existing customer base. Positive free cash flow gives Tesla additional flexibility, but one quarter does not establish the long-term profitability of emerging businesses that remain under development. Musk has also discussed a future combination involving SpaceX, although no definitive merger agreement has been announced. Investors must therefore distinguish completed financial results from proposed transactions and longer-term technology ambitions. Tesla’s challenge is to show that its newer products can become dependable commercial businesses rather than remaining primarily research and development programs.

Why It Matters: Tesla’s earnings highlight the financial tension between sustaining its electric vehicle business and investing heavily in autonomous driving, robotics, and AI.

Source: Associated Press.

Barclays Plans 5,000 Job Cuts as AI and Technology Automation Reshape Banking Operations

British bank Barclays is cutting approximately 5,000 jobs, or around 5% of its workforce, as it pursues organizational changes and technology-driven efficiency improvements. The company disclosed the reductions alongside its third-quarter earnings, according to Business Insider reporting published Thursday. Barclays linked the restructuring to organizational simplification and efficiency gains associated with artificial intelligence. Approximately 60% of the affected positions are expected to be in support functions, including software development and other internal operations. Group finance director Anna Cross said AI is increasing employee productivity and contributing to structural changes in the bank’s cost base. The announcement does not establish that AI systems will directly perform every task previously handled by the affected employees, since the reductions also involve broader organizational changes.

The decision adds to a growing debate over how AI will affect professional employment. Financial institutions are experimenting with technology that can analyze documents, prepare reports, assist software developers, answer customer questions, and automate administrative processes. Those capabilities may improve productivity while reducing demand for certain forms of repetitive work. However, regulated financial services still require human responsibility for compliance, complex customer decisions, and risk management. Banks must also verify that automated systems operate reliably and do not introduce security or legal problems. For workers, the shift could increase demand for technical oversight, specialized judgment, and the ability to manage AI-assisted workflows. For enterprise software startups, Barclays’ announcement shows why customers increasingly want products that promise measurable reductions in operating costs rather than incremental improvements in convenience.

Why It Matters: Barclays’ restructuring provides another example of how AI adoption is influencing workforce decisions as large enterprises pursue productivity improvements.

Source: Business Insider.

Northwestern Physicists Trap Single Antimatter Particle for Months in Quantum Computing Research Breakthrough

Researchers at Northwestern University have demonstrated unusually precise quantum control over a single positron, the antimatter counterpart of an electron, while keeping the particle isolated for months. The achievement was described in research published in Physical Review Letters and highlighted by Phys.org on Wednesday. The team, led by physicist Gerald Gabrielse, used specialized electric and magnetic fields to hold the particle in a near-perfect vacuum while cooling it to extremely low temperatures. Under those conditions, the positron behaves in ways that researchers can describe through carefully controlled quantum energy states. The scientists call the resulting experimental system a one-positron quantum cyclotron. The demonstration represents progress in fundamental particle physics rather than the creation of a commercial quantum computer.

The experiment could eventually support more precise comparisons between matter and antimatter. According to established physical theories, electrons and positrons should exhibit closely related properties, with certain differences determined by their opposite electric charges. Researchers want to test those predictions with increasing accuracy because any unexpected discrepancy could reveal limitations in existing theories. The question is connected to one of physics’ unresolved mysteries: why the observable universe contains much more matter than antimatter. Maintaining control over a single positron for extended periods creates opportunities to measure its properties under unusually stable conditions. However, the latest experiment does not resolve that mystery or demonstrate a new computing architecture ready for commercial use. Its significance lies in improving experimental methods that may support future discoveries. The research illustrates how advances in quantum control can contribute to fundamental science even when immediate applications remain uncertain.

Why It Matters: Northwestern’s demonstration could enable more precise tests of fundamental physics and improve researchers’ ability to study the differences between matter and antimatter.

Source: Phys.org / Northwestern University.

Hugging Face Releases Open Carbon-X Model That Lets Researchers Chat With DNA

Hugging Face released Carbon-X on Thursday, a 14-billion-parameter open model that integrates DNA sequences with natural language so users can ask questions about genetic variants, regulatory regions, and biological effects in plain English. Developed with collaborators at Zhongguancun Academy in China and the University of Naples Federico II, the model was mid-trained on hundreds of billions of tokens mixing genomic data from mammals, plants, fungi, and other organisms with English text. It uses a hybrid tokenizer that processes DNA in six-letter groups and can run locally on suitable hardware under an Apache 2.0 license.

Carbon-X is intended to help researchers design drugs and genetic therapies by predicting whether mutations are benign or disease-linked and by generating candidate sequences. On internal DNA benchmarks it approaches the performance of much larger closed models such as GPT-6.1 Sol while operating at a fraction of the inference cost, and it can serve as a specialist sub-agent for other systems. Limitations include a context window of roughly 50,000 base pairs, the absence of wet-lab validation of its predictions, and deliberate exclusion of viral sequences.

Why It Matters: Open, conversational genomic models lower the barrier for academic and startup researchers outside frontier labs while raising questions about validation and dual-use risks.

Source: WIRED.

Anthropic Launches Claude Duet Proactive AI Agent App Powered by Opus 5.5

Anthropic opened a research preview of Claude Duet, a standalone iOS, Android, and web app in which its Claude agent works proactively rather than waiting for prompts. Aimed at professionals and small businesses, Duet delivers morning briefings drawn from connected apps, monitors ongoing tasks, drafts materials, and surfaces items needing approval; it is built on the Claude Opus 5.5 model and can tap more than 6,000 connectors in the Claude Marketplace, including Gmail and Slack.

The agent can also initiate purchases via Stripe Link or Shopify Shop Pay but requires explicit user approval for sensitive actions. Unlike more personal assistants, Anthropic positions Duet as less suited to errands such as booking tickets and emphasizes restraint on credit-card or identity requests by default. Access is initially limited to US Pro and Max subscribers via a waitlist and shares existing usage limits. The launch places Anthropic in a crowded field of agentic tools from OpenAI, Meta, and others that aim to complete multi-step work with reduced human oversight.

Why It Matters: Proactive agents that act across enterprise tools accelerate the shift from chat interfaces to autonomous workflow software, intensifying competition and safety scrutiny.

Source: TechCrunch.

PHP2Shell Attacks Surge After Emergency PHP Security Patches as Researchers Warn of Widespread Scanning

Cybersecurity researchers are observing widespread attempts to exploit two critical PHP vulnerabilities following the release of emergency security updates on October 21. The flaws, tracked as CVE-2026-109618 and CVE-2026-109619, have been collectively described as PHP2Shell. According to SecurityWeek reporting published Thursday, Amazon detected suspicious attempts targeting the weaknesses as early as October 17, several days before the vulnerabilities were publicly disclosed. Additional proof-of-concept exploit code appeared after information about the flaws circulated online. Security companies subsequently observed large numbers of automated requests probing potentially vulnerable servers. However, researchers warned that many publicly circulating exploits do not reliably achieve remote code execution, and widespread scanning should not be confused with confirmed mass compromise.

The latest developments represent a significant change from Wednesday’s initial warnings, when administrators were still preparing for the release of fixes. Updated PHP versions are now available, including 8.5.12, 8.4.27, 8.3.36, and 8.2.35. The U.S. Cybersecurity and Infrastructure Security Agency has urged organizations to update affected installations promptly. This is especially important for businesses operating multiple websites, older applications, or systems managed by different hosting providers. Vulnerable PHP installations may exist on forgotten servers or internal applications that administrators rarely inspect. Sites running unsupported PHP branches may require additional remediation or migration work. Although firewall protections can reduce exposure, they are not a substitute for fixing the underlying vulnerabilities. The incident demonstrates how quickly attackers can begin testing newly disclosed weaknesses in foundational web technologies.

Why It Matters: The surge in PHP2Shell probing makes rapid patching and accurate inventories of PHP installations increasingly important for website owners and hosting providers.

Source: SecurityWeek.

Google DeepMind Researchers Resign Over Military AI Contracts With US and Israeli Forces

At least two Google DeepMind researchers have resigned in recent weeks over the company’s expanding military AI contracts. London-based research engineer Sofia Liguori left this week after three years, telling the BBC she felt ashamed that the company was pushing AI systems into military operations and that she had lost faith in its adherence to stated ethical principles. Researcher Andreas Kirsch departed earlier in October, citing insufficient governance and visibility into how models are used in classified Pentagon settings.

Google has signed agreements permitting AI use for lawful government purposes, with human oversight required for certain autonomous or surveillance applications, and maintains that public-sector contracts undergo review against its AI principles. The departures add to longer-running internal tensions over Project Nimbus and other defense work, highlighting the difficulty frontier labs face in retaining talent while pursuing large government contracts.

Why It Matters: Public resignations at leading AI labs over military applications increase pressure on companies to demonstrate meaningful control over classified deployments.

Source: BBC.

Ukrainian Drone Strikes Damage Russian Data Centers, Disrupting Yandex Services and AI Research Infrastructure

Ukrainian drone attacks have damaged several Russian data centers, disrupting digital services and potentially setting back artificial intelligence research, according to a new Associated Press investigation. The attacks occurred between October 8 and October 18 and affected infrastructure associated with major Russian technology companies, including Yandex, Sberbank, and VKontakte. Satellite imagery examined by the AP showed extensive damage at a Yandex-linked facility in Sasovo, in Russia’s Ryazan region. The site reportedly housed important computing equipment used in research and commercial services. The investigation found that the strikes affected infrastructure supporting online banking, e-commerce, and other digital activities. However, the full extent of the damage and the time required to restore services remain uncertain.

The findings illustrate how data centers have become strategically important infrastructure during modern conflicts. Facilities that once primarily supported websites and business software now also operate AI systems, financial services, and communications platforms used by large populations. Damage to such installations can therefore affect both commercial activity and technological development. Replacement may be particularly difficult when specialized processors, networking components, and supporting equipment are subject to international export restrictions. Experts interviewed by the AP warned that restoring some affected computing systems could be time-consuming. The attacks also raise broader questions about protecting civilian digital infrastructure during wartime. For governments and cloud operators worldwide, the investigation demonstrates the risks of concentrating essential computing capacity in a limited number of physical locations without sufficient operational redundancy.

Why It Matters: The damage to Russian data centers demonstrates how physical attacks on computing infrastructure can disrupt digital economies and potentially undermine long-term AI development.

Source: Associated Press.

AI Research Startup Flapping Airplanes Raises $1 Billion at Reported $6 Billion Valuation to Build More Efficient Models

Flapping Airplanes, an artificial intelligence research startup founded approximately a year ago, has raised $1 billion in Series A funding at a reported $6 billion valuation. Index Ventures and Lux Capital led the financing, with participation from investors including Sequoia Capital, GV, and Kleiner Perkins. The company previously raised approximately $180 million in seed financing at a valuation of roughly $1.5 billion. Its new funding will support research into whether advanced AI models can be developed more efficiently than the increasingly expensive systems built by the largest frontier laboratories. Unlike companies selling established enterprise applications, Flapping Airplanes is primarily a research organization and has not yet demonstrated a widely adopted commercial product. Its valuation therefore reflects investors’ expectations about future breakthroughs rather than an established record of revenue or profitability.

The investment highlights a growing divide in AI development strategies. Major laboratories are committing enormous amounts of capital to larger models, additional training data, and extensive computing infrastructure. Flapping Airplanes is exploring whether improvements in learning methods and model design could produce competitive results with fewer resources. That approach could have substantial commercial implications if it reduces the cost of training or operating sophisticated systems. However, advances demonstrated under controlled research conditions do not necessarily translate into production systems that can compete across many different tasks. The company will need to demonstrate meaningful improvements, attract technical talent, and eventually identify a sustainable business model. For the venture capital industry, the round illustrates continued willingness to finance ambitious foundational research despite growing scrutiny of AI valuations and infrastructure spending.

Why It Matters: Flapping Airplanes’ financing shows that investors see major commercial potential in reducing the computational cost of advanced AI rather than simply building larger models.

Source: TechStartups via Flapping Airplanes.

Meta Reports $64.6 Billion in Q3 Revenue as AI Spending Pushes Annual Capital Budget Toward $145 Billion

Meta reported third-quarter revenue of $64.6 billion on Wednesday, up 26% from a year earlier, as advertising growth continued to fund its expanding artificial intelligence ambitions. The Facebook and Instagram parent generated $9.86 billion in net income, though its results included a $9.89 billion legal charge tied to an earlier agreement with U.S. state attorneys general over youth-related social media litigation. Operating expenses increased substantially during the quarter, while capital expenditures, including finance lease principal payments, reached $34.91 billion. Meta now expects full-year 2026 capital spending of $135 billion to $145 billion, compared with its previous projection of $130 billion to $145 billion. The company also reported negative free cash flow of $2.12 billion for the quarter.

The results illustrate the unusual economics of the world’s largest AI investors. Meta operates an enormously profitable advertising business, giving it resources that independent AI laboratories must often obtain through external financing. However, building frontier models requires continued investment in processors, servers, networking, and electricity-intensive data centers. Meta is also funding consumer AI products and the research needed to compete with other major developers. Its quarterly advertising performance suggests the company’s established business remains commercially strong, but it does not show that its newer AI investments will produce comparable financial returns. Management expects substantial spending to continue as it expands computing capacity and develops additional products. Investors will increasingly examine whether improvements in advertising performance, customer engagement, and new AI revenue streams justify that expenditure.

Why It Matters: Meta’s earnings demonstrate how profitable technology platforms are financing enormous AI infrastructure investments even as rising costs pressure cash flow.

Source: Meta Platforms.

Swedish Legal AI Startup Legora Raises $550 Million at $8.8 Billion Valuation as Annual Recurring Revenue Tops $200 Million

Swedish legal technology startup Legora announced Wednesday that it has raised $550 million in Series E financing at an $8.8 billion post-money valuation. Atreides Management led the round, with participation from investors including Baron Capital, Capital Group, Fidelity Management & Research Company, and several existing backers. Legora develops AI software that helps legal professionals review documents, research legal questions, prepare drafts, and coordinate work across matters. The company says its annual recurring revenue has exceeded $200 million, up from $100 million less than six months earlier. It also reports that more than 140,000 legal professionals use its technology across more than 2,400 law firms and in-house legal departments. Those operating figures are company-reported and have not been independently audited for this report.

The financing reflects the growing market for specialized AI applications that address expensive professional workflows. Legal teams spend substantial time analyzing contracts, examining evidence, comparing documents, and preparing routine materials. AI tools can reduce some of that administrative work, potentially letting lawyers focus on client advice, negotiations, and professional judgment. However, legal applications also face demanding requirements involving confidentiality, accuracy, and responsibility for the information they produce. An incorrect legal citation or mishandled confidential document can create consequences far beyond an ordinary software error. Legora’s commercial expansion suggests that law firms are increasingly willing to adopt AI systems, but continued growth will depend on demonstrated reliability and customer retention. Its valuation also shows how quickly enterprise AI companies can attract substantial capital when investors believe they are establishing a strong position within a specialized industry.

Why It Matters: Legora’s latest round highlights investor demand for AI software that automates specialized professional work while meeting enterprise customers’ security and accuracy requirements.

Source: Legora.

Nokia Reports 17% Network Infrastructure Sales Growth as AI Data Centers Drive Demand for Optical Networking

Nokia reported a 17% increase in third-quarter network infrastructure sales as demand from AI data centers continued to support its optical networking business. The Finnish technology company said revenue from AI and cloud customers more than doubled from a year earlier, reflecting growing demand for equipment used to connect large computing facilities. Group net sales reached approximately €5.15 billion. Chief Executive Justin Hotard highlighted continued strength in infrastructure serving cloud providers and AI-related customers. Nokia is also progressing with the integration of Infinera, the optical networking company it acquired to expand its position in high-capacity communications technology. Management said the integration remains on track and expects to complete key elements ahead of its original schedule.

The results demonstrate that the AI infrastructure boom extends far beyond semiconductor manufacturers. Training advanced models and serving large numbers of users require enormous volumes of data to move between processors, servers, and geographically distributed facilities. This creates demand for fiber-optic networking equipment that can transmit information quickly and reliably. Nokia is attempting to strengthen its position in this market as telecommunications companies and cloud operators modernize their infrastructure. However, growth driven by AI data centers does not eliminate challenges elsewhere in the telecommunications equipment industry, where traditional operator spending can fluctuate. Component availability, pricing, and competition will also influence future results. For investors and startups, Nokia’s performance illustrates how companies supplying foundational networking technology can benefit from AI adoption without developing language models themselves.

Why It Matters: Nokia’s results show how the expansion of AI computing is creating substantial opportunities for companies supplying the networking infrastructure required to connect data centers.

Source: The Wall Street Journal.

Google Expands Chrome Security With AI Scam Detection and New Defenses Against ClickFix Malware Attacks

Google is expanding Chrome’s security protections to identify malicious websites that attempt to trick users into executing harmful commands on their own computers. The company is introducing additional defenses against a social engineering technique known as ClickFix, in which fraudulent websites present fake errors or verification instructions and persuade visitors to copy and run dangerous commands. Google is also introducing a new category of suspicious-site warnings alongside its existing alerts for websites confidently identified as malicious. The changes are part of Chrome’s broader Safe Browsing system. For users who enable Enhanced Protection, Google is expanding server-side Gemini analysis to help identify fraudulent pages, including fake antivirus and browser-update websites.

The developments reflect a major shift in cybercrime techniques. Rather than exploiting a browser vulnerability directly, attackers increasingly try to persuade users to take actions that bypass standard technical safeguards. ClickFix campaigns can be effective because they disguise malicious instructions as routine troubleshooting steps. Google’s system examines signals associated with potentially fraudulent websites and combines AI analysis with reputation information and other security indicators. The company says it handles page content sent for AI analysis temporarily and does not retain it or use it for model training. Those assurances matter because inspecting website content through remote systems can create additional privacy questions. Google also says its Enhanced Protection mode generates substantially more warnings than its standard offering, although that metric should not be interpreted as a direct measurement of prevented infections. The broader significance is that browser security is expanding beyond detecting malicious files toward identifying manipulative instructions presented to users.

Why It Matters: Chrome’s new protections could reduce the effectiveness of phishing and social engineering campaigns that depend on persuading users to execute malicious commands.

Source: SecurityWeek.

IBM Reports $17.1 Billion in Q3 Revenue as Enterprise Software Growth Supports AI Strategy

IBM reported third-quarter revenue of $17.1 billion Wednesday, up 5% from a year earlier, as stronger software sales helped offset slower growth in other parts of the business. Software revenue increased 8%, while consulting revenue remained approximately flat on a reported basis. Infrastructure revenue rose 4%. The company reported non-GAAP operating earnings of $2.95 per share and generated $7.8 billion in free cash flow during the first nine months of 2026. IBM is continuing to position its software, consulting, and infrastructure businesses around enterprise AI adoption, including automation, data management, and hybrid cloud technology. However, not all of the company’s revenue growth can be attributed specifically to AI, since its portfolio includes established products and services serving many different business needs

The results provide a useful comparison with the economics of younger AI companies. IBM operates mature businesses with established customers and recurring commercial relationships, allowing it to introduce AI capabilities through existing enterprise software and consulting arrangements. That distribution advantage can reduce some customer acquisition challenges while creating opportunities to expand software usage. However, large organizations often require extensive integration work before new AI systems deliver meaningful benefits. Businesses must ensure that AI applications work with existing databases, comply with internal security requirements, and produce measurable improvements in productivity or decision-making. IBM’s relatively modest overall revenue growth illustrates that AI adoption does not automatically transform the financial performance of every established technology supplier. Instead, successful implementation may occur gradually as customers modernize systems and identify applications that justify the cost.

Why It Matters: IBM’s results demonstrate how established enterprise technology companies are incorporating AI into profitable software and services businesses without relying exclusively on frontier-model development.

Source: IBM.

OPPO Launches Find X10 Flagship Phones With 200MP Cameras, Advanced Chips, and AI Features

Chinese smartphone manufacturer OPPO has introduced its Find X10 flagship lineup internationally, expanding competition in the premium Android market with high-resolution cameras and new AI software. The range includes the Find X10 and Find X10 Pro Max, both emphasizing advanced mobile photography developed with Hasselblad. According to reporting published Wednesday by Spanish business newspaper Cinco Días, the standard model includes a dual 200-megapixel camera configuration, while the Pro Max features three 200-megapixel sensors. The higher-end model also incorporates a large silicon-carbon battery rated at 7,150 mAh, with support for fast wired and wireless charging. OPPO is introducing ColorOS 17, based on Android 17, alongside the new devices. Spanish pricing begins at approximately €1,399 for the standard model and €1,799 for the Pro Max.

The launch demonstrates how premium smartphone manufacturers are competing across hardware, computational photography, battery technology, and AI-assisted software. High-resolution sensors can capture substantial image detail, but actual photographic performance depends on lens quality, image processing, lighting conditions, and software calibration. OPPO’s battery technology also reflects efforts to increase energy capacity without dramatically increasing device size. The company is adding more AI capabilities to its operating system, although manufacturers across the industry continue experimenting with which features consumers find valuable enough to use regularly. OPPO faces strong competition from Apple, Samsung, Xiaomi, and other Android manufacturers. Its ability to gain market share will depend on pricing, distribution, camera performance, software reliability, and long-term support. Manufacturer specifications provide a useful starting point, but independent testing will be needed to show how the devices perform in everyday use.

Why It Matters: OPPO’s new phones highlight continued innovation in consumer hardware as smartphone companies combine advanced cameras, battery technology, and AI software to compete in the premium market.

Source: Cinco Días.

Palantir Says NATO’s Maven AI System Has No Remote Kill Switch as Europe Questions Dependence on U.S. Technology

Palantir says neither the company nor the United States government can remotely disable the Maven Smart System used by NATO’s military headquarters, responding to growing European concerns about dependence on American defense technology. Louis Mosley, Palantir’s executive vice president for the United Kingdom and Europe, told Reuters that NATO controls the system’s operation and that no remote mechanism allows Palantir to shut it down. NATO acquired the technology in 2025 to support military planning and intelligence analysis. The statement follows renewed concern among European officials after disruptions involving Ukraine’s access to military-grade satellite communications. Different parties have offered conflicting explanations for those disruptions, and Palantir’s technical assurances have not been independently verified through publicly available system testing.

The issue highlights a broader challenge for governments adopting commercial AI technologies for sensitive national security applications. Advanced software can help analyze large datasets, identify patterns, and support operational planning, but customers also need confidence that systems will remain available during political disputes, technical failures, or disruptions to supporting infrastructure. NATO’s adoption of Maven does not mean every member country uses the platform for its own armed forces. Some European governments continue evaluating alternatives and emphasizing greater technological independence. For defense technology startups, the debate creates opportunities to develop systems with locally controlled infrastructure, transparent operational safeguards, and reduced reliance on external suppliers. However, sovereign control involves more than software access. Governments may also depend on foreign processors, cloud infrastructure, technical support, and future security updates. Palantir’s assurances address one concern while leaving those broader dependencies relevant.

Why It Matters: The debate over NATO’s AI systems highlights growing demand for defense technology that governments can operate independently of external suppliers and political decisions.

Source: Reuters.