Top Tech News Today, May 20, 2026

It’s Wednesday, May 20, 2026, and the AI arms race just kicked into overdrive. Google dropped its most ambitious agentic models yet at I/O, Big Tech is locked in a frantic scramble for chips, power, and talent, regulators blinked on rules that were supposed to bite this summer, and fresh signals from China, Nasdaq debuts, and venture flows show exactly where the next wave of winners (and pressure points) is forming. Here are the 15 stories making waves today.

Today’s biggest tech stories show the same pattern playing out across the globe: Google is pushing Gemini into nearly every corner of its ecosystem, Nvidia is tightening its grip on AI infrastructure, Saudi Arabia is pouring billions into compute, and regulators are starting to treat advanced AI like a national security issue.

Here are the top global technology news stories you need to know today.

Technology News Today

Meta begins first wave of 8,000 layoffs in AI efficiency push and restructuring

Meta initiated the first round of its planned 10% workforce reduction today, cutting roughly 8,000 roles as part of an efficiency drive tied to heavy AI spending. The company had already reassigned about 7,000 employees to new AI-focused groups in recent days, reorganizing teams to prioritize model development and infrastructure. Managers adapted structures for AI initiatives, with further cuts expected later in 2026. The moves follow Meta’s broader $725 billion industry-wide AI capex wave alongside peers like Google and Tesla.

The layoffs reflect a sector-wide pattern where Big Tech trims non-core roles to redirect resources toward AI infrastructure and talent. Insiders note the reassignments aim to retain key expertise amid the transformation, but the timing underscores the tension between growth ambitions and operational streamlining. Meta, which employed over 78,000 at the end of 2025, had signaled the changes weeks earlier.

Why It Matters: Meta’s simultaneous layoffs and AI reallocations highlight how Big Tech is reshaping its workforce to dominate the AI race, influencing talent strategies and efficiency benchmarks across the startup and tech ecosystem.

Source: Bloomberg.

Google I/O 2026 puts Gemini at the center of Google’s AI future

Google used I/O 2026 to unveil a broad wave of Gemini-powered products across Search, Android, Workspace, YouTube, shopping, and AI hardware. The company introduced Gemini 3.5 Flash, Gemini Omni, Gemini Spark, AI coding tools, AI-generated interfaces, and new partnerships for Android XR glasses.

The message was clear: Google is no longer treating AI as a layer on top of its products. It is rebuilding the product stack around Gemini. That matters because Google still controls massive distribution across search, mobile, email, browsers, video, and cloud. If it can make AI useful across those surfaces, it could slow the rise of AI-native challengers.

Why It Matters: Google is turning its ecosystem into a full-stack AI platform.

Source: TechStartups via Google I/O.

Big tech layoffs and job cuts increasingly tied to AI efficiency drives

AP News highlighted a wave of layoffs at companies including Cisco, Block, and others, where executives explicitly cited AI-driven efficiencies as a factor in workforce reductions. The trend mirrors broader industry shifts toward automation and the prioritization of AI.

This pattern illustrates how AI is not only creating new roles but also displacing others, prompting discussions on reskilling and the net impact on employment in tech-heavy sectors. It affects startup hiring strategies as talent pools adjust.

Why It Matters: Linking job cuts to AI underscores the technology’s dual role in driving efficiency and workforce transformation, shaping labor policies and talent strategies across the global tech ecosystem.

Source: AP News.

Google Search gets its biggest AI overhaul yet

Google is changing Search with a new Gemini 3.5 Flash-powered search box, longer queries, file attachments, AI Overviews, AI Mode, and agent-style tools that can monitor information or act on a user’s behalf.

This is one of the clearest signs yet that the old “ten blue links” model is fading. For publishers, startups, advertisers, and SEO teams, the shift raises a major question: how much traffic will still flow to the open web when Google can answer more queries inside its own AI interface?

Why It Matters: Google’s AI Search shift could reshape discovery, publishing, and startup customer acquisition.

Source: The Verge.

Young workers voice growing skepticism toward AI bots in workplace

Reuters surveyed sentiment showing younger employees increasingly skeptical or resistant to AI agents replacing routine tasks, citing concerns over job quality, creativity, and oversight. The backlash emerges even as adoption accelerates.

This cultural pushback could slow enterprise rollouts and prompt companies to prioritize human-AI collaboration models. It underscores the importance of addressing workforce perceptions alongside technological advances.

Why It Matters: Rising youth skepticism toward AI bots highlights a critical barrier to adoption, prompting companies to focus on ethical integration and upskilling to maintain productivity gains.

Source: Reuters.

Google DeepMind hires Contextual AI researchers in licensing deal

Google DeepMind struck a licensing agreement with Contextual AI and is hiring more than 20 of the startup’s researchers, including co-founder and CEO Douwe Kiela. The deal is reportedly valued between $80 million and $90 million.

The move fits a growing pattern in AI: Big Tech companies are gaining access to talent and technology without traditional acquisitions. These deals can move faster than full buyouts, but they also attract more regulatory scrutiny because they may avoid merger review.

Why It Matters: AI talent is now so valuable that licensing-plus-hiring deals are becoming acquisition alternatives.

Source: Reuters.

CEOs distribute AI tokens as performance incentives amid massive tech spend

Fortune reported that executives at leading firms are increasingly using AI-related tokens or equity as compensation to align teams with AI priorities, even as companies justify record infrastructure outlays. This practice emerges as Big Tech collectively commits hundreds of billions to AI, with some leaders cutting non-AI roles to fund it.

The trend reflects a cultural shift in which AI fluency is becoming a core metric for advancement, helping to retain talent in a competitive market. It also raises questions about valuation and accountability as firms balance hype with tangible returns on massive investments.

Why It Matters: The rise of AI tokens as pay signals a deeper integration of AI into corporate culture and incentives, influencing how startups and enterprises attract and motivate tech talent.

Source: Fortune.

Anthropic’s Mythos AI model hacking risks deemed overstated by experts

Reuters analysis found that concerns over unfettered hacking enabled by Anthropic’s new Mythos AI model were exaggerated, based on early testing and safeguards in place. The model’s capabilities sparked debate over the security implications of advanced AI systems.
The assessment calms some regulatory and enterprise fears while underscoring the need for robust testing as models grow more powerful. It informs ongoing policy discussions around AI safety without stifling innovation.

Why It Matters: Clarifying risks around models like Mythos helps balance security concerns with rapid AI advancement, guiding responsible deployment by Big Tech and startups.

Source: Reuters.

Microsoft secures $9.7 billion data center deal with IREN for Nvidia chips and capacity

Microsoft signed a major $9.7 billion partnership with IREN, gaining access to Nvidia chips and expanded data center capacity to support its growing AI workload demands. The deal also includes an expansion of its collaboration with Lambda to bolster cloud infrastructure for Azure AI services. This comes amid surging needs for compute power as Microsoft integrates advanced models across its ecosystem.

The agreement underscores the intense scramble for AI infrastructure amid power and chip constraints, with hyperscalers locking in long-term capacity to avoid bottlenecks. It positions Microsoft to scale training and inference more aggressively against rivals, while highlighting the booming market for specialized data center operators. Analysts see it as part of a broader $400 billion annual spend by Big Tech on AI facilities.

Why It Matters: Microsoft’s massive IREN partnership intensifies the AI infrastructure arms race, driving demand for Nvidia hardware and reshaping how cloud giants secure competitive edges in compute and energy resources.

Source: Cloudcomputing-news.net.

Saudi AI company Humain taps Goldman for $5.3B data center financing

Saudi Arabia’s Humain has selected Goldman Sachs to advise on financing for a data center project that could cost at least 20 billion riyals, or about $5.33 billion. The PIF-backed company wants to build data centers and acquire GPUs as part of a broader AI infrastructure push.

The project highlights the Middle East’s growing role in the AI infrastructure race. Cheap energy, state capital, and national diversification plans are turning Saudi Arabia, the UAE, and Qatar into serious contenders for hyperscale AI compute.

Why It Matters: AI infrastructure is becoming a geopolitical competition, not just a Silicon Valley spending race.

Source: Reuters.

Nvidia’s $90B AI deal spree deepens its grip on the compute economy

Nvidia has committed roughly $90 billion across deals and investments over the past 16 months, backing more than 145 companies across AI developers, cloud providers, infrastructure firms, and hardware suppliers.

The strategy gives Nvidia more than chip sales. It gives the company influence across the entire AI supply chain, from cloud capacity to networking, model builders, and future demand. But it also raises scrutiny because Nvidia can act as supplier, investor, customer, and ecosystem gatekeeper at the same time.

Why It Matters: Nvidia is building an AI economy around itself.

Source: Financial Times.

Blackstone and Google launch $25 billion AI infrastructure venture to challenge Nvidia’s dominance

Google and Blackstone are backing a new AI cloud venture designed to expand the market for Google’s custom Tensor Processing Units. The project includes an initial $5 billion investment from Blackstone and plans for 500 megawatts of data center capacity in 2027.

The move is a direct challenge to Nvidia’s dominance in AI infrastructure. Google already has a strong chip story with TPUs, but the missing piece has been broader market access. Pairing with Blackstone gives Google a financial and infrastructure partner that can help scale the model.

Why It Matters: Google is trying to turn TPUs into a broader AI cloud alternative to Nvidia GPUs.

Source: TechStartups via BlackStone, Financial Times.

Venture funding roundup shows AI infrastructure and agent startups leading deals

Recent venture capital activity, per market trackers, concentrated on AI infrastructure, agents, and frontier applications, with several nine-figure rounds closed in the past day. Funding reflected continued optimism despite broader economic signals.

This momentum sustains startup innovation in high-priority areas, enabling faster scaling of technologies that Big Tech may later acquire or integrate. It also highlights regional strengths in the U.S. and Asia.

Why It Matters: Strong AI-focused venture flows signal sustained investor confidence, fueling the pipeline of breakthroughs that will define the next phase of the tech ecosystem.

Source: TechStartups.com (via funding roundup aggregation).

AI startup Decart raises $300M to make switching chips easier

Decart raised $300 million in a round led by Radical Ventures, with Nvidia participating, valuing the startup near $4 billion. Its Decart Optimization Stack helps AI developers move workloads across chips from Nvidia, Amazon, Google, and others.

That makes Decart especially interesting. Nvidia is investing in a company whose software could make AI labs less dependent on Nvidia hardware. The bigger signal is that AI infrastructure is moving beyond raw chips into portability, optimization, and cost control.

Why It Matters: Startups that reduce AI infrastructure lock-in are becoming highly valuable.

Source: Wall Street Journal.

Trump AI order may seek early government access to advanced models

The Trump administration is preparing an AI executive order that would encourage frontier AI developers to notify the government and critical infrastructure providers before releasing advanced models. The draft reportedly includes a 90-day early warning framework.

The proposal reflects growing concern that powerful AI systems could create new cyber risks before regulators understand their capabilities. For AI labs, it could also mark a shift toward closer government oversight before release.

Why It Matters: AI safety is moving from voluntary pledges to national-security-style oversight.

Source: Axios.

Microsoft disrupts fake certificate service used by ransomware gangs

Microsoft dismantled infrastructure tied to Fox Tempest, a cybercriminal service that sold fake code-signing certificates to ransomware groups. The operation seized domains, websites, and Azure resources linked to the network.

Fake certificates help malware look legitimate, making them valuable to attackers trying to bypass security tools. Microsoft said the group produced more than 1,000 fake certificates and ran hundreds of Azure accounts.

Why It Matters: Code-signing abuse remains a key weapon in ransomware operators’ arsenal.

Source: Axios.

New Shai-Hulud malware wave hits more than 600 npm packages

A new Shai-Hulud supply-chain campaign pushed more than 600 malicious packages to npm, targeting the JavaScript ecosystem. Security researchers said the attack affected the AntV ecosystem and related packages, with credential theft and cloud secret harvesting among the risks.

The incident is another warning for startups and enterprise teams that developer tooling is now a major attack surface. One compromised maintainer account can ripple through thousands of apps, CI/CD systems, and cloud environments.

Why It Matters: Software supply-chain attacks are becoming faster, broader, and harder to contain.

Source: BleepingComputer, Snyk.

Amazon’s secret ‘Titus’ project shows Nvidia’s hold on AI infrastructure

Amazon’s internal “Titus” project reportedly involves redesigning AWS data centers to support Nvidia’s advanced GPU systems, including upcoming GB200 racks.

That is notable because Amazon has been pushing its own Trainium chips as an alternative. The project suggests AWS still needs to align closely with Nvidia’s roadmap to meet customer demand for high-end AI compute.

Why It Matters: Even hyperscalers building their own chips remain deeply tied to Nvidia.

Source: Business Insider.

Primer raises $100M as payments infrastructure attracts fresh funding

London-based payments infrastructure startup Primer raised $100 million, with Peak XV joining the round. Primer helps companies simplify complex payment systems across markets and providers.

The funding shows investor interest remains strong in infrastructure startups that sit beneath consumer-facing fintech. Payments remain fragmented globally, and companies want more control over routing, checkout, fraud, and provider redundancy.

Why It Matters: Fintech infrastructure continues to attract capital when it solves painful operational problems.

Source: Economic Times.

GlobalFoundries backs deep tech AI startup fund

GlobalFoundries invested in Playground Global’s $475 million Fund IV through its GF Accelerates venture program. The fund targets deep tech startups working on AI hardware, chip efficiency, energy use, and next-generation semiconductor tools.

The move shows how chip manufacturers are trying to stay close to early-stage innovation. As AI workloads strain existing hardware, startups working on materials, lithography, memory, and power efficiency could become strategically important.

Why It Matters: The AI chip race is creating new opportunities for deep tech startups.

Source: Times Union.

Google’s Gemini Spark pushes AI agents closer to everyday work

Google unveiled Gemini Spark, a personal AI agent designed to draft emails, assemble documents, monitor inboxes, and eventually complete purchases even when a user’s devices are inactive.

The product points to where consumer AI is heading: from chatbots that answer questions to agents that perform tasks. The challenge will be trust, permissions, privacy, and reliability, especially when agents begin acting across apps and money-related workflows.

Why It Matters: The next AI battleground is task completion, not just conversation.

Source: VentureBeat.

Nature papers show AI agents moving deeper into scientific discovery

Nature published research on AI systems designed to assist scientific discovery, including Robin, a multi-agent system that can automate hypothesis generation and data analysis for experimental biology. A related Nature release said the systems are meant to assist researchers rather than replace them.

The work is important because AI is moving beyond summarizing papers or generating lab notes. If these systems can help design experiments, interpret results, and generate new hypotheses, they could change how biotech, pharma, and university labs operate.

Why It Matters: AI agents are starting to enter the core workflow of scientific research.

Source: Nature.