OpenAI Files for IPO, Anthropic Nears First Profit, Google Remakes Search — AI News Briefing

🗞️ AI News Briefing — May 24, 2026 (06:00 CST)

Top 7 Stories

1. OpenAI Confidentially Files for IPO — The Trillion-Dollar Countdown Begins

OpenAI is preparing to file confidentially for an initial public offering as soon as this week, according to multiple reports from CNBC, Bloomberg, and The New York Times. The filing, expected imminently, marks the culmination of years of transformation from a nonprofit research lab into one of the most valuable private companies in the world. Sam Altman’s company is reportedly aiming for a valuation that could approach $1 trillion, which would make it one of the largest technology IPOs in history.

The IPO filing comes at a pivotal moment for the AI industry. OpenAI faces intensifying competition from Anthropic, which has been rapidly closing the gap in enterprise adoption, and Google, which just unveiled its Gemini Omni model at I/O 2026. The public filing will force OpenAI to disclose detailed financial information for the first time, including the true scale of its revenue, losses, and the staggering compute costs associated with training and running its models.

For investors and the broader tech industry, OpenAI’s IPO represents a critical test of whether the market will reward AI companies at the valuations their founders and backers have been demanding. The filing arrives alongside SpaceX’s own IPO prospectus, creating what analysts are calling “mega-IPO season” — a period that could redefine how public markets value artificial intelligence businesses. The coming weeks will determine whether AI hype translates into sustained public market enthusiasm.

2. Anthropic Nears First Profitable Quarter — Revenue Hits $10.9 Billion in Q2

Anthropic is on track to report its first profitable quarter, with revenue expected to reach $10.9 billion in the second quarter of 2026, according to CNBC and the Wall Street Journal. This milestone represents an extraordinary trajectory for a company founded in 2021 and signals that Anthropic’s focus on enterprise AI and safety-first positioning is paying off commercially. The company recently surpassed OpenAI in enterprise adoption, according to DigiTimes, as businesses increasingly favor Anthropic’s Claude models for their reliability and lower hallucination rates.

In a striking display of the AI infrastructure arms race, Anthropic has agreed to pay SpaceX $1.25 billion per month for computing power, Reuters reported. This deal underscores how AI companies are locking in access to compute resources at unprecedented scales — and how SpaceX is diversifying its revenue beyond rockets and Starlink into the AI infrastructure market. Forbes noted that Anthropic and OpenAI are now taking “opposite paths to AI profitability,” with Anthropic pursuing margin-focused growth while OpenAI continues to burn cash in pursuit of market dominance.

Anthropic’s imminent profitability, combined with its growing enterprise footprint, positions the company as a formidable competitor ahead of its own expected IPO. With Andrej Karpathy recently joining the pre-training team, Anthropic is simultaneously strengthening its technical bench and its commercial momentum — a combination that has made it the #1 ranking on CNBC’s 2026 Disruptor 50 list.

3. Andrej Karpathy Joins Anthropic — OpenAI Co-Founder Makes the Leap

In one of the most significant personnel moves in the AI industry this year, OpenAI co-founder Andrej Karpathy has joined Anthropic’s pre-training team, Axios and TechCrunch confirmed. Karpathy, who previously led AI development at Tesla and was one of OpenAI’s earliest employees, brings deep expertise in large-scale model training, computer vision, and autonomous systems. His move from OpenAI to Anthropic sends a powerful signal about the shifting competitive dynamics between the two leading AI labs.

Karpathy’s departure from OpenAI to join a direct competitor is rare at the co-founder level and speaks to Anthropic’s growing appeal as a technical destination. At Tesla, Karpathy built the self-driving AI stack from the ground up; at OpenAI, he contributed to foundational research on language models and neural networks. His expertise in pre-training — the computationally intensive process of teaching AI models on massive datasets before fine-tuning — is particularly valuable as Anthropic scales its next generation of Claude models.

The hire also reflects a broader trend of talent redistribution in the AI industry. As Anthropic, Google, and other well-funded competitors scale their research operations, they are attracting top researchers from OpenAI and other labs. For Anthropic, Karpathy’s arrival strengthens its pre-training capabilities at a critical moment — just as the company is approaching profitability and preparing for a potential IPO. For OpenAI, it’s a reminder that even founding-team loyalty can’t withstand the gravitational pull of competing labs with deeper pockets and fresh momentum.

4. Google I/O 2026: Gemini Omni and the Biggest Search Transformation in 25 Years

Google used its I/O 2026 developer conference to announce what it’s calling “a new era for AI Search,” unveiling Gemini Omni — its most advanced AI model to date — alongside a suite of personal AI agents that fundamentally change how users interact with the search engine. According to CNBC, The Verge, and Wired, these updates represent the most significant change to Google’s search box in 25 years, as the company shifts from returning links to delivering agentic, task-completing AI experiences.

The new AI agents can perform multi-step tasks on behalf of users — from planning trips to completing online forms to conducting research across multiple sources — all within the Google ecosystem. Google is pitching this as an “AI agent ecosystem” for consumers, according to TechCrunch, in an effort to keep pace with OpenAI’s ChatGPT and Anthropic’s Claude, both of which have gained significant consumer traction. Sundar Pichai acknowledged the anxiety people feel about AI’s rapid advancement in a New York Times interview, while simultaneously pushing Google’s most aggressive AI product rollout ever.

The implications for the broader internet economy are profound. Google’s shift toward AI-driven search results means fewer clicks to external websites, a trend that has already begun affecting traffic to news sites and publishers. At the same time, Google is upgrading its AI search advertising platform, giving marketers new ways to reach users within AI-generated responses. For advertisers, developers, and content creators, Google I/O 2026 marked the beginning of a fundamentally different relationship between search and the web.

5. AI Tokenmaxxing Crisis: Tech Giants Curb AI Use as Agentic AI Costs Spiral

A growing cost crisis is gripping the tech industry as companies discover that internal AI usage — particularly agentic AI workflows — is consuming vastly more tokens and costing far more than anticipated. Tom’s Hardware reported that agentic AI agents can consume up to 1,000 times more tokens than standard AI interactions, leading to what employees have dubbed “tokenmaxxing” — gaming AI usage metrics to hit arbitrary corporate targets. Microsoft, Meta, and Amazon have all begun curbing internal AI deployments as costs spiral out of control.

Fortune reported that Microsoft’s internal data is revealing AI’s “real cost problem”: in many cases, using AI for routine tasks is more expensive than paying human employees to do the same work. The phenomenon has led to Microsoft canceling some internal Claude licenses and reassessing its Copilot strategy, according to India Today. At Amazon, HR built leaderboards to encourage AI adoption, but workers are now exploiting these systems to maximize token consumption without generating proportional value, TechRadar reported.

This cost reckoning is forcing a fundamental reassessment of AI’s ROI across the enterprise sector. The initial assumption that AI would universally reduce costs is giving way to a more nuanced reality: AI can be transformative for specific high-value use cases, but blanket deployment across all workflows often destroys value. The tokenmaxxing phenomenon — where employees are incentivized to use AI excessively to meet corporate KPIs — has created a perverse feedback loop that is burning through AI budgets without delivering commensurate productivity gains. Industry analysts expect this to lead to a more selective, ROI-driven approach to enterprise AI adoption in the second half of 2026.

6. California’s Landmark AI Workforce Order — Newsom Signs First-of-Its-Kind Executive Action

California Governor Gavin Newsom has signed a first-of-its-kind executive order aimed at preparing the state’s workers and businesses for AI-driven workplace disruption, the California State Portal confirmed. The order comes in the wake of growing concerns about AI-related job displacement and follows reports that tech layoffs have already surpassed 100,000 in 2026 as companies redirect budgets toward AI investments. The executive order directs state agencies to develop a data-driven framework focused on protecting workers and supporting small businesses through the AI transition.

CalMatters reported that the order was issued specifically “after AI layoffs,” as Newsom ordered state government to find ways to ease the pain for displaced workers. The framework includes provisions for workforce retraining programs, early-warning systems for AI-driven layoffs, and support mechanisms for small businesses navigating AI adoption. Better Markets praised the order as a “good start,” noting its data-driven approach to monitoring AI’s impact on jobs and small businesses.

However, the order has faced criticism from labor advocates who argue that the protections don’t go far enough. CBS8 reported that critics say the order’s safeguards “fall short” of what’s needed to protect California’s workforce, particularly in industries like tech, media, and customer service where AI displacement is already visible. The order represents the most significant state-level AI workforce policy in the United States and could serve as a model for other states grappling with similar challenges. As AI continues to reshape the labor market, California’s experiment will be closely watched by policymakers nationwide.

7. Trump Delays AI Oversight Executive Order — Industry Pushback halts Last-Minute Signing

President Trump delayed the signing of an executive order on AI oversight just hours before it was scheduled to be finalized, Politico and The Washington Post reported. The order, which would have established new federal oversight mechanisms for AI model development and deployment, was pulled back after AI industry leaders — including White House AI czar David Sacks — raised concerns about its potential impact on U.S. competitiveness. Scott Bessent, the Treasury Secretary, has also been raising alarms about AI policy, though implementation delays continue to mount.

Politico obtained the unsigned order and published its full text, revealing provisions that would have created new regulatory frameworks for AI safety testing, model evaluation, and deployment monitoring. The delay highlights the ongoing tension between the administration’s desire to regulate AI and the tech industry’s push for minimal government interference. This follows earlier reports from The Guardian that Trump had blasted an AI company for standing firm on safety guardrails that the U.S. military wanted lifted.

The postponement leaves a regulatory vacuum at a time when AI capabilities are advancing faster than any existing policy framework can address. With the EU’s AI Act already in effect and China advancing its own AI governance framework, the U.S. delay raises questions about America’s role in setting global AI standards. The order’s fate remains uncertain, and its delay — coming alongside California’s proactive workforce order — illustrates the fragmented and often contradictory nature of U.S. AI policy in 2026.

📊 Trend Watch

🔭 What to Watch

• Dell Q1 Earnings (May 28) — Dell stock has surged 16% to an all-time high ahead of its fiscal Q1 report. With HPE and Super Micro also rallying, the earnings call will be a key barometer of AI server demand and infrastructure spending momentum heading into the second half of 2026.

• OpenAI IPO Filing Details — The confidential S-1 filing will reveal OpenAI’s actual revenue, losses, and growth trajectory for the first time. The numbers will either validate or challenge the trillion-dollar valuation narrative and could impact valuations across the entire AI sector.

• Anthropic’s First Profitable Quarter — If Anthropic delivers on its projected Q2 profitability with $10.9B in revenue, it will become the first major AI lab to prove sustainable unit economics — a milestone that could reshape investor expectations for the entire industry and accelerate its own IPO timeline.