Nvidia Smashes Records with $68B Quarter, Samsung Unveils 'Agentic AI' Galaxy S26, Perplexity Launches Computer, Anthropic Drops Safety Pledge
Nvidia crushes Wall Street expectations with $68.1 billion in Q4 revenue and begins shipping Vera Rubin samples. Samsung unveils the Galaxy S26 series with "truly agentic AI" at Unpacked in San Francisco. Perplexity launches Computer, an autonomous agent that orchestrates 19 AI models. Anthropic quietly loosens its core safety promise amid the Pentagon standoff. Economists push back on Citrini's viral doomsday report. Anthropic acquires computer-use startup Vercept. OpenAI names Arvind KC as Chief People Officer.
1. Nvidia Smashes Wall Street Expectations with Record $68.1 Billion Quarter
Nvidia delivered the most anticipated earnings report in tech on Wednesday evening, and it didn't disappoint. The AI chipmaker posted record fiscal Q4 2026 revenue of $68.1 billion — beating analyst expectations by roughly $3 billion — with adjusted earnings of $1.62 per share versus the $1.53 consensus. Revenue soared 73% year-over-year, driven almost entirely by the company's data center business, which climbed 75% to $62.3 billion.
But the real headline came from the forward-looking guidance: Nvidia projected fiscal Q1 2027 revenue of $78.0 billion (plus or minus 2%), signaling that demand for AI compute continues to accelerate rather than plateau. CNBC reported the stock was up slightly in late trading, extending its rally for the year.
On the same call, Nvidia announced it has begun delivering samples of its next-generation Vera Rubin platform to select customers. Tom's Hardware detailed the specs: the Vera Rubin superchip pairs two Rubin GPUs (each with 288 GB of HBM4 memory) with an 88-core Vera CPU, totaling 17,000 components across a single board. The system promises 10x better performance per watt than Grace Blackwell, with 1.3 million components sourced from over 80 suppliers across 20 countries.
"Customers are racing to invest in AI. The age of AI is in full steam." — Jensen Huang, Nvidia CEO, via Fortune
Vera Rubin is expected to ship in the second half of 2026, and TipRanks noted the platform is positioned as Nvidia's answer to AMD's growing threat from its MI540 series — which just landed the massive $60 billion Meta deal earlier this week. Citi reiterated its buy rating on Nvidia, calling the stock likely to "outperform in the second half of 2026" as hyperscaler AI capex accelerates.
Source: CNBC, Fortune, Investopedia, Tom's Hardware
The $78 billion guidance for next quarter is the real story here — it tells us AI infrastructure spending is still in its exponential growth phase, not approaching a plateau. For enterprise leaders planning AI infrastructure, the Vera Rubin specs (10x performance per watt over Blackwell) mean that waiting 6-12 months before making major GPU commitments could yield dramatically better price-performance ratios. But don't wait passively — start architecture planning now so you can deploy rapidly when Vera Rubin ships. The companies that win will be those with infrastructure blueprints ready to execute, not those starting from scratch when better hardware arrives.
2. Samsung Unveils Galaxy S26 Series: 'The Beginning of Truly Agentic AI'
Samsung held its Galaxy Unpacked event in San Francisco on Wednesday, unveiling the Galaxy S26 series with a bold positioning statement: this is "the beginning of truly agentic AI" on mobile devices. The Galaxy S26 Ultra features a 39% more powerful NPU (Neural Processing Unit) for faster on-device AI processing and a 19% faster CPU, with Samsung framing the improvements around immediacy rather than theoretical capability.
Samsung's Global Newsroom highlighted several AI-first features: circling a celebrity's outfit in a photo generates curated shopping results to recreate the look; the phone integrates with Bixby, Google Gemini, and Perplexity simultaneously; and a new multimodal AI camera understands context in real-time. Samsung Business Insights emphasized the "agentic" framing — the phone doesn't just respond to commands, it proactively anticipates needs and takes action across apps.
Google also made announcements tied to the launch. Google's blog revealed that the Galaxy S26 will be the first device to feature Gemini-powered scam detection directly in Samsung's Phone app, using on-device intelligence to analyze calls in real-time and warn users of potential fraud without sending data to the cloud.
"The goal isn't theoretical capability, but immediacy. Every interaction should feel like the phone already knows what you need." — Samsung, via Samsung Business Insights
Source: Samsung Global Newsroom, Google Blog, Samsung Business Insights
Samsung's "agentic AI" language is significant — it signals that the smartphone industry is moving beyond the "AI assistant" paradigm toward autonomous agents that act on your behalf across apps. For businesses with mobile-first customer experiences, this changes the game: your app needs to be accessible not just to humans tapping buttons, but to AI agents navigating on a user's behalf. If your mobile experience isn't API-friendly or structured for agent interaction, you're about to become invisible to a growing segment of users who let their phone's AI do the browsing and buying for them.
3. Perplexity Launches 'Computer' — An Autonomous Agent Orchestrating 19 AI Models
Perplexity AI introduced its most ambitious product yet on Tuesday: Computer, an autonomous agent platform designed to complete complex multi-step projects with minimal human supervision. The product unifies research, design, coding, deployment, and project management into a single system, orchestrating 19 different AI models simultaneously to tackle end-to-end workflows.
Semafor described Computer as "a super agent that can create a website, write a report, or generate a dataset," while The Hindu noted it represents Perplexity's expansion from an answer engine to "a comprehensive productivity platform." Implicator reported that Perplexity is positioning Computer as a direct response to the failures of earlier autonomous agents, though the company doesn't name competitors specifically.
"Introducing Perplexity Computer. Computer unifies every current AI capability into one system. It can research, design, code, deploy, and manage any project end-to-end." — Perplexity, via X/Twitter
PYMNTS framed the launch in the context of a broader shift toward autonomous agents, noting that "agents are increasingly operating through application interfaces rather than human-centric screens, performing work that used to require human oversight." The multi-model orchestration approach — using specialized models for different subtasks rather than relying on a single monolithic model — represents a growing architectural trend in the agent space.
Source: Semafor, PYMNTS, The Hindu, Implicator
The 19-model orchestration architecture is the most interesting technical detail here. Rather than building one model to rule them all, Perplexity is betting that specialized models working in concert will outperform generalists — essentially applying microservices thinking to AI. For enterprise architects evaluating agent platforms, this is a critical design question: do you want a single-vendor monolithic agent (like Claude Cowork) or a multi-model orchestrator? The answer depends on your tolerance for complexity versus your need for best-in-class performance at each subtask. We're watching this space closely because the winning architecture pattern will define enterprise AI for the next 3-5 years.
4. Anthropic Quietly Loosens Core Safety Promise Amid Pentagon Standoff
In a move that sent shockwaves through the AI safety community, Anthropic on Wednesday quietly updated its Responsible Scaling Policy — the foundational document that has distinguished it as the most safety-conscious major AI lab. Marketplace reported that Anthropic acknowledged it had "failed to persuade other companies to take its stricter safety approach" and was loosening its commitments accordingly.
CNN called it a moment where Anthropic "ditches its core safety promise," noting the timing — right in the middle of the company's showdown with the Pentagon over military AI access. The policy change means Anthropic may continue building more advanced AI systems even if competitors aren't matching its safety benchmarks, reversing a prior commitment to slow down development if the industry didn't follow suit.
The move also came on the same day that NBC News reported Anthropic had offered the Pentagon the ability to use Claude for missile defense applications — a significant concession from a company that previously maintained strict guardrails against military use cases. Lawfare published a detailed legal analysis of whether the Defense Production Act could actually compel Anthropic to comply, noting the contractual guardrails that currently "prohibit applications such as autonomous weapons and mass surveillance."
Source: CNN, Marketplace, NBC News, Lawfare
This is a pivotal day for AI governance. Anthropic's entire brand was built on the premise that it would prioritize safety even at the cost of competitive position. That promise is now being walked back under dual pressure — commercial competition from OpenAI's aggressive enterprise push, and government coercion from the Pentagon. For enterprise buyers who chose Anthropic specifically for its safety credentials, this is a moment to reassess. Your AI governance framework should not depend on any single vendor's voluntary commitments — it needs to be your own, enforced through contracts, audits, and technical controls that you own and operate.
5. Economists Push Back on Citrini's Viral AI Doomsday Scenario
The Citrini Research report that rattled markets earlier this week — painting a dystopian 2028 scenario where AI agents trigger mass unemployment and economic collapse — faced a coordinated backlash from economists and institutional investors on Wednesday. The Economist published a pointed rebuttal titled "A viral research note on AI gets its economics wrong," arguing that the scenario fundamentally misunderstands how labor markets adjust to technological disruption.
Bloomberg reported that the pushback was global, with investors and economists from multiple continents criticizing the report's methodology. The New York Times noted that while many analysts questioned the conclusions, the report's viral spread had already done its damage — Monday's Dow dropped 820 points (1.7%) and the S&P 500 fell 1%, though markets rebounded Tuesday.
Inc. described the piece as "a thought exercise in financial history, from the future" — a creative framing that resonated with anxious traders even as professional economists dismissed its core assumptions. The episode highlighted a growing tension: the market is simultaneously pricing AI as transformative and terrifying, creating volatility from narrative shifts as much as fundamentals.
Source: The Economist, Bloomberg, New York Times, Inc.
The Citrini episode is a case study in narrative risk. A Substack post from a small firm moved billions in market cap because it tapped into a genuine, widespread anxiety that the market hadn't priced properly. For business leaders, the lesson isn't whether Citrini's specific scenario is right or wrong — it's that AI disruption narratives now have the power to move markets in ways that affect your cost of capital, your customers' spending behavior, and your talent pipeline. Your communications strategy needs a proactive AI narrative that positions your company as a beneficiary of AI transformation, not a victim of it. Silence leaves space for doomsday scenarios to fill.
6. Anthropic Acquires Computer-Use Startup Vercept
In a deal that flew somewhat under the radar amid the Pentagon drama, Anthropic on Wednesday announced it has acquired Vercept, an AI startup specializing in computer-use agents — software that can navigate desktop and web applications autonomously. TechCrunch reported that Vercept has "deep roots to some of the biggest names in Seattle's tech scene," and the acquisition came after Meta reportedly poached one of Vercept's co-founders.
The acquisition signals Anthropic's aggressive push into the agentic AI market. Computer-use capabilities — where AI agents can click buttons, fill forms, navigate websites, and operate software on behalf of users — have become a critical competitive frontier. OpenAI's Operator, Perplexity's Computer (launched the same day), and Google's Mariner all compete in this space. By acquiring Vercept's specialized team and technology, Anthropic is bolstering its Claude Cowork platform's ability to interact with enterprise software at the UI level, not just through APIs.
Source: TechCrunch
The Vercept acquisition is strategically smart: computer-use agents that can interact with legacy software UIs solve the "last mile" integration problem that plagues enterprise AI deployments. Many businesses run on applications that don't have APIs — they have screens that humans click through. An AI agent that can navigate those screens is the bridge between the AI-native future and the legacy-software present. For IT leaders, this means your AI strategy shouldn't wait for every vendor to ship API integrations. Computer-use agents can automate workflows across applications that were never designed for automation. Start identifying those manual, screen-clicking workflows now.
7. OpenAI Appoints Arvind KC as Chief People Officer Amid IPO Preparations
OpenAI announced Tuesday that it has appointed Arvind KC as its new Chief People Officer, effective February 24. KC joins from Roblox, where he served as Chief People and Systems Officer, and he becomes the latest Indian-origin executive in OpenAI's C-suite, joining Vijaye Raji and others. The hire was reported by The Hindu Business Line, The Economic Times, and Moneycontrol.
In his new role, KC will oversee OpenAI's global people strategy — hiring, onboarding, organizational development, and the systems and policies that support a company experiencing hypergrowth. The appointment comes at a critical juncture: OpenAI is in the midst of a $100 billion funding round, is reportedly preparing for a potential public listing later this year, and is scaling rapidly to compete with Anthropic's enterprise push and Google's deepening AI integration across its product suite.
Source: The Hindu Business Line, The Economic Times, Pulse2
A Chief People Officer hire from Roblox — a company that scaled through a chaotic hypergrowth phase and went public — is a clear signal that OpenAI is in pre-IPO mode. For enterprise customers evaluating OpenAI partnerships, a public OpenAI means more transparency (SEC filings, quarterly earnings calls) but also more pressure to show revenue growth, which could affect pricing and product strategy. For talent leaders, the appointment signals that the AI talent war is entering its next phase: the companies that win will be those that can attract, retain, and develop talent at scale, not just those with the best models.
🔍 Why It Matters for Business
Wednesday's headlines reveal an AI industry operating on multiple fronts simultaneously. Nvidia's record earnings confirm that AI infrastructure spending is accelerating, not plateauing. Samsung's "agentic AI" phone launch signals that autonomous agents are going mainstream on consumer devices. Perplexity's multi-model Computer shows the agent architecture debate is far from settled. And Anthropic's safety policy retreat demonstrates that even the most principled AI companies bend under competitive and government pressure.
The throughline: AI is no longer coming — it's here, it's accelerating, and the rules governing it are being rewritten in real-time. Businesses that wait for clarity before acting will find themselves too far behind to catch up.
Need help navigating AI for your business?
Our team turns these developments into actionable strategy — from architecture to deployment.
Contact SEN-X →