This Month's Latest Tech News in the US - February 28th 2026 Edition

In This Update

• February at a Glance: AI Infrastructure Becomes the Main Story
• Nvidia hits $5 trillion and the AI infrastructure arms race
• Agentic AI: Google, startups, and the next wave of autonomous agents
• AI-driven layoffs and the hot hiring market
• Practical upskilling: tech-to-trades and Nucamp bootcamps
• OpenAI’s $110B round, Plaid liquidity, and the IPO comeback
• AI regulation and 'Responsible Scaling': guardrails versus regulatory
• Energy strategy: data centers, nuclear, and sodium-ion batteries
• How AI is changing work, products, and everyday tech
• Market signals and next steps: events to watch and what to do

Nvidia hits $5 trillion and the AI infrastructure arms race

Nvidia spent February rewriting Wall Street record books, becoming the first company to reach a roughly $5.03 trillion market cap as its stock traded near $207.16, capping a about 39% gain over the past year. At the same time, the chipmaker committed $1 billion to Nokia, a move aimed at shoring up the networking backbone that AI-era data centers and telecom operators will rely on, as highlighted in recent coverage of major tech players by StartUs Insights.

Far from a meme-stock spike, Nvidia’s run reflected its status as the de facto tollbooth of the AI age: it controls much of the high-end GPU supply that hyperscalers, startups, and enterprises need. CEO Jensen Huang projected that between $3 trillion and $4 trillion could be spent on AI infrastructure by 2030, with AI companies themselves shouldering much of the bill. Analysts at outlets like Yahoo Finance grouped Nvidia alongside other high-growth U.S. tech names benefiting from that shift.

Those forecasts were already materializing in a wave of “Stargate”-scale projects: multi-billion-dollar joint ventures between OpenAI, SoftBank, Oracle, and others to build gargantuan AI data centers and connectivity hubs. Former President Donald Trump reportedly described one such initiative as the “largest AI infrastructure project in history,” while OpenAI CEO Sam Altman framed it in epochal terms.

“I think this will be the most important project of this era.” - Sam Altman, CEO, OpenAI

The result is an AI infrastructure arms race that now extends from chips into telecom, construction, and energy. For free-market advocates, the concern is not excess private capital but whether permitting, power access, and local land-use battles can keep up. If red tape slows grid upgrades and data center approvals, incumbents with the deepest pockets and best lobbying operations may entrench their lead while smaller players struggle to plug in at all.

Agentic AI: Google, startups, and the next wave of autonomous agents

From chatbots to AI colleagues

While infrastructure giants grabbed the biggest headlines, February also marked a clear shift from static chatbots to agentic AI - systems that can plan, act, and hand off work. Google published research on an “adaptive delegation framework,” designed to let AI agents route tasks to humans or other agents when confidence is low, a step toward making autonomous systems robust enough for real workflows rather than demo scripts.

Healthcare, finance, and humanoid robots

Agent deployments accelerated in tightly regulated industries. In healthcare, providers such as ElevenLabs and Retell AI expanded use of voice and workflow agents for medical billing and patient intake, according to a February roundup from Peterson Technology Partners. In finance, AI agents now account for roughly 20% of new partnerships, as banks and fintechs look to automate onboarding, customer support, and compliance-heavy back-office processes.

On the hardware side, Figure AI’s humanoid robots became a symbol of how fast capital is moving into embodied agents: the startup reached a valuation of about $39 billion in just three years, the fastest unicorn trajectory in robotics history, according to startup trackers. That kind of bet suggests investors see agents not just as software utilities but as a new labor platform that can plug directly into warehouses, factories, and retail.

New skills, old anxieties

For developers, the opportunity sat in the glue code: orchestration frameworks, API integrations, monitoring, and safety tooling that keep agents from going off the rails. Enterprise leaders, however, were more focused on whether customers and employees would trust these systems. In its 2026 Technology Trends release, consultancy CapTech argued that winning organizations would pair AI with transparent design and change management, a theme highlighted in coverage of the CapTech report.

“AI is deeply linked to how organizations grow and compete, but real progress now depends on helping people understand and trust AI.” - Brian Bischoff, CTO, CapTech

AI-driven layoffs and the hot hiring market

Layoffs remained a defining feature of February’s tech news cycle, even as headline valuations hit new highs. Autodesk cut about 7% of its workforce, or roughly 1,000 jobs, explicitly to refocus spending on its cloud platform and AI initiatives. Pinterest announced plans to eliminate just under 15% of staff as part of “transformation initiatives” around AI-powered products, while insurance software vendor Sapiens shed roughly 540 roles across the U.S. and India, according to tallies compiled by The American Bazaar.

Executives framed these cuts as necessary “efficiency” moves to free up capital for infrastructure, data, and automation. For employees, especially in operations, support, and middle management, the message was more blunt: white-collar roles once assumed to be insulated from automation are now squarely in scope. A Global News segment asked whether U.S. AI layoffs should be a “wake-up call” for other advanced economies, underscoring how quickly corporate cost-cutting logic can travel across borders.

The hiring data, however, told a more nuanced story. LinkedIn figures summarized in the February 2026 Tech Job Market Trends report from Zero To Mastery showed a 26.6% month-over-month surge in AI engineer postings. Demand was especially strong for roles that sit close to AI infrastructure and production deployments, including:

• AI and machine learning engineers
• Data engineers and data platform specialists
• DevOps and cloud reliability engineers
• Security engineers in AI-heavy environments

Inside enterprises, an Alteryx study cited in the same ecosystem found that fewer than 1 in 4 AI pilots make it into production, largely because of data quality and integration problems rather than model performance. That gap highlights a classic creative destruction pattern: generic roles and legacy layers shrink, while specialists who can wrangle data pipelines, harden cloud systems, and ship AI safely into production see their bargaining power rise.

Practical upskilling: tech-to-trades and Nucamp bootcamps

As AI hiring concentrated around specialized roles in infrastructure, data, and security, a parallel “tech-to-trades” trend emerged in February. Harvard Business Review described workers pivoting toward more hands-on, “AI-proof” skilled trades to regain control over their careers, while others looked for pragmatic ways to stay in the tech game without taking on six-figure tuition or pausing full-time work.

That search for market-driven reskilling options put affordable bootcamps under a brighter spotlight. Among them, Nucamp’s affordable coding bootcamps stood out for mid-career workers trying to move closer to the AI stack without going back to college. The school offers several focused tracks aligned with where hiring is strongest:

Nucamp positioned itself as one of the most affordable bootcamps in the industry, with programs designed around a part-time commitment of roughly 10-20 hours per week. Small cohorts and live instruction aimed to differentiate it from massive, video-only platforms, while bundled career services promised practical help with portfolios and interviews rather than abstract credentials.

For workers caught between AI-driven layoffs and a hot but narrow hiring market, this kind of targeted, low-cost training reflects a broader free-market response: instead of waiting for government retraining schemes to materialize, individuals can buy specific skills - back-end Python, DevOps, cybersecurity, or AI product-building - at a predictable price and timeline, then test that investment directly in the job market.

OpenAI’s $110B round, Plaid liquidity, and the IPO comeback

Money moved at historic scale around AI and fintech in February. OpenAI closed what Crunchbase described as the largest venture deal ever, pushing the company to an estimated valuation near $840 billion and underscoring investor conviction that foundation models are a long-term platform, not a passing fad. Fintech infrastructure provider Plaid, meanwhile, completed a major tender offer at about an $8 billion valuation, giving early employees and investors liquidity without the scrutiny of a full IPO.

Those late-stage deals unfolded alongside a broader reopening of U.S. public markets. After several sluggish years, the IPO window “roared back” in February for AI-native and adjacent software companies, with bankers reporting stronger order books and less price sensitivity for firms that could show real revenue tied to automation, data, or cloud infrastructure. That shift was reflected in high-growth public names as well, where investors rewarded companies converting AI narratives into top-line expansion.

Analysts highlighted a handful of standouts. Marker Therapeutics posted revenue growth of roughly 62.86%, while Palantir Technologies reported about 26.25% growth, according to breakdowns of U.S. high-growth tech stocks. Looking ahead, forecasts suggested Veracyte could reach around $582 million in revenue by 2026, outpacing the broader market and signaling that not all AI-related upside is concentrated in the biggest consumer platforms.

Still, the enthusiasm came with a caution label. Deutsche Bank strategist Jim Reid told Fortune’s analysis of the AI-driven software selloff that “nobody truly knows who the long-term winners and losers of this extraordinary technology will be,” warning that sharp rotations between favored AI names could wipe out paper gains quickly. For retail investors and employees with concentrated equity, that uncertainty reinforced a familiar lesson: broad, adaptable skills may be a safer long-term asset than any single AI stock, no matter how hot the current cycle looks.

AI regulation and 'Responsible Scaling': guardrails versus regulatory

Regulators spent February moving from speeches to statutes, turning AI governance into a concrete line item rather than a talking point. A market analysis on Stocktwits warned that “2026 is coming for Big Tech and AI with a battery of new laws”, flagging proposals that would regulate model training data, biometric tracking, and platform liability for AI-generated content. For large platforms, that looked like a new compliance cost of doing business; for smaller startups and open-source projects, it raised questions about whether they could even afford to play.

In parallel, major labs leaned into self-imposed guardrails. Companies such as Anthropic promoted “Responsible Scaling” protocols for their frontier models, designed to prevent systems from assisting with sensitive tasks like biological weapons development or sophisticated cyberattacks. These frameworks typically bundle red-teaming, staged capability evaluations, and kill switches, aiming to reassure both policymakers and enterprise customers that safety isn’t an afterthought.

The political flashpoint came when President Trump ordered U.S. agencies to stop using Anthropic’s technology after the company refused to relax its red lines. Anthropic reiterated that it would not support mass domestic surveillance or fully autonomous weapons and insisted that “humans should remain in the loop for high-stakes automated decisions,” according to reporting on the Pentagon AI deals by Radio New Zealand. House Democratic leader Hakeem Jeffries praised the stance, calling broader surveillance proposals an “invasion of privacy scheme.”

• Civil-liberties advocates welcomed clearer limits on state surveillance and autonomous weapons.
• National-security officials argued aggressive controls could slow U.S. defense innovation.
• Startup founders worried that only the richest firms would be able to navigate overlapping safety and privacy rules.

From a pro-innovation perspective, the tension is less about whether guardrails are needed and more about how they are written. Narrow, risk-based rules can protect citizens while leaving room for competition; sprawling, one-size-fits-all mandates risk hard-coding today’s giants into permanent gatekeepers, raising barriers to entry for new labs, entrepreneurs, and independent researchers who lack an army of lawyers but may have the next breakthrough model.

Energy strategy: data centers, nuclear, and sodium-ion batteries

Power strategy quietly became one of February’s most important tech stories. As AI data centers scaled up, operators increasingly treated electricity less as background “IT overhead” and more as a core input on par with GPUs, signing long-term deals with utilities and exploring everything from advanced renewables to next-generation nuclear to secure predictable, politically acceptable supply.

Industry reporting described large AI projects negotiating multi-decade power contracts and scouting regions with surplus generation capacity, even as local communities raised concerns about water use, noise, and grid strain. In parallel, emerging storage technologies moved closer to the spotlight. MIT Technology Review’s “10 Breakthrough Technologies 2026” list named sodium-ion batteries as a key innovation with the potential to decentralize the grid and reduce dependence on lithium, alongside “generative coding” tools that could reshape how software is built, according to MIT’s emerging tech coverage.

Sodium-ion systems rely on far more abundant materials than lithium-ion cells, making them attractive for large-scale stationary storage rather than smartphones. For AI infrastructure, that could translate into cheaper local buffers that smooth the sharp peaks in data center demand, making it easier to integrate intermittent solar and wind without constant reliance on gas peaker plants. In effect, batteries and smarter software become part of the AI stack, not just adjunct utilities.

Yet the speed of build-out continues to run into 20th-century rules. Nuclear siting processes that stretch over many years, transmission projects stalled by permitting fights, and local zoning disputes over new substations all risk favoring incumbents with the legal budgets and lobbying muscle to navigate them. Commentators in roundups like SingularityHub’s late-February tech digest noted that the same private capital driving AI could accelerate cleaner, more resilient grids - if regulators modernize frameworks to match the new reality.

For engineers and founders, that makes basic fluency in energy - how data centers are powered, where batteries fit, and what local rules allow - an increasingly valuable skill. As AI turns into heavy industry, those who can bridge software, hardware, and power policy will be better positioned to influence where the next wave of infrastructure gets built and who benefits from it.

How AI is changing work, products, and everyday tech

On the ground, February’s AI story showed up less in boardroom speeches and more in day-to-day workflows. In Deloitte’s 2026 Tech Trends report, one CIO admitted that “the time it takes us to study a new technology now exceeds that technology’s relevance window,” capturing how quickly generative models, copilots, and agents were hitting employees’ screens compared with traditional rollout cycles.

Research discussed in Harvard Business Review’s look at nine trends shaping work in 2026 introduced two ideas now worrying managers. “Workslop” describes productivity lost when staff must fix low-quality AI outputs, turning promised efficiency into digital busywork. “Digital doppelgängers” refers to AI-generated versions of employees used in support, training, or sales - raising new questions about consent, performance metrics, and who benefits when a synthetic clone handles the night shift.

Consumer tech offered a softer face of the same shift. AI-driven wellness devices such as the Ambient Dreamie bedside companion drew early praise for improving sleep routines, reflecting a broader “AI for health and habits” trend. On the phone front, Google’s Pixel 10a continued the company’s playbook of modest hardware bumps paired with smarter on-device AI at an aggressive price point; reviewers at outlets like Engadget’s technology news and reviews hub emphasized value over specs for the mid-range device.

Mobility and urban life saw visible change as well, with Waymo expanding its robotaxi operations to 10 U.S. cities by late February. For residents in those markets, autonomous vehicles shifted from sci-fi demo to mundane option for cross-town trips. Inside enterprises, meanwhile, many AI projects remained stuck in pilot mode, hampered by messy data, brittle integrations, and unclear accountability. That disconnect between flashy demos and durable deployment underscored a central lesson of the month: the real leverage lies not just in building smarter models, but in redesigning work, products, and customer journeys around them in ways that people actually trust.

Market signals and next steps: events to watch and what to do

February’s tape told a clear story: markets continued to reward companies that could turn AI into measurable growth, not just headline-grabbing demos. Mega-cap platforms with profitable AI lines, like Alphabet - whose shares were up roughly 65% over the past year - outperformed broader indexes, while newly listed, AI-heavy software names saw strong demand so long as they showed real revenue and not just user metrics.

Off the trading floor, a cluster of industry events reinforced the same themes. Enterprise trend reports and accessibility-focused showcases at major trade shows all pointed toward three priorities: building trust in AI systems, treating data as a product, and keeping humans meaningfully in the loop. Deloitte’s 2026 Tech Trends analysis, for instance, argued that organizations integrating AI with human-centered design, rather than bolting on tools, were best positioned to capture value, a point explored in detail on Deloitte Insights’ technology management series.

Heading into March, several fault lines looked set to define the next phase:

• Whether agentic AI moves from pilots into mission-critical workflows, creating new roles in AI operations, safety review, and orchestration.
• How far Washington pushes on AI safety, surveillance, and content rules - and whether compliance costs disproportionately burden smaller firms.
• How local backlash over power, water, and land shapes where new data centers and grid upgrades can actually be built.

For workers and builders, those signals point toward action rather than paralysis. The practical playbook is to move closer to the AI value chain - data, infrastructure, security, and applied product work - while diversifying away from roles tied to legacy stacks or manual reporting. That often means using private-sector options such as focused bootcamps, industry certificates, and on-the-job experimentation to upskill quickly, instead of waiting for large employers or government programs to dictate the pace of change.