AI's Competitive Scramble Intensifies as Market Bubble Fears, Job Disruption Narratives, and Geopolitical Rivalries Converge

Anthropic Dominance, Pentagon Politics, and a Reshuffled AI Race Leaderboard

Perscient's semantic signature tracking language asserting that Anthropic or Claude leads the artificial intelligence competition stands at 319, the highest current value among all tracked signatures, and strengthened by 10 points over the past week. Anthropic's Claude Cowork platform, which rattled Wall Street last month with its initial legal plugin, received a wider release on February 24 with new enterprise connectors and plugins, spanning investment banking, human resources, and beyond. The company's annualized revenue has climbed to $14 billion, fueled in part by Claude Code, its AI coding tool, which alone has reached $2.5 billion in annualized revenue, and business subscriptions have quadrupled since the start of the year. Anthropic's most recent $30 billion funding round, led by Coatue and Singapore sovereign wealth fund GIC with participation from D.E. Shaw Ventures, Dragoneer, Founders Fund, and others, has pushed the startup's valuation to $380 billion.

Yet just as Anthropic consolidates its lead, the competitive picture is shifting. Our semantic signature tracking language asserting that Grok or xAI leads the AI competition rose by 26 points to 83, one of the largest weekly moves across all tracked signatures. The U.S. military formally signed an agreement to deploy xAI's Grok model inside classified systems for intelligence analysis, decision support, and weapons development tasks. Until now, Anthropic's Claude had been the only AI model permitted inside those systems, making this a direct incursion into Anthropic's most privileged domain.

The Pentagon backdrop is tense. Defense Secretary Pete Hegseth has summoned Anthropic CEO Dario Amodei to the Pentagon for what sources describe as a likely heated discussion over the company's restrictions on military use of Claude. According to the AP, Pentagon officials are warning that they could designate Anthropic a supply chain risk or invoke the Defense Production Act if the company does not yield. As the New York Times reported, Amodei has long maintained that AI needs strict limits to prevent it from wrecking economies or starting wars, and reportedly did not budge on fully autonomous military targeting operations or domestic surveillance. The most commercially dominant AI company is being told by its biggest government customer to loosen restrictions or face exclusion, while its chief rival enters through the door Anthropic's own principles may have left ajar.

Meanwhile, our signature tracking language asserting that Google or Gemini leads the AI race declined by 10 points to 83, while the equivalent signature for OpenAI weakened by 5 points to -19, well below the long-term mean. Narrative attention is migrating from these traditional frontrunners toward Anthropic and, increasingly, xAI.

Geopolitical narratives around AI competition also intensified. Perscient's semantic signature tracking language predicting Chinese AI dominance rose by 44 points to -8, the second-largest weekly increase across all signatures, still below average but closing ground rapidly. Our signature tracking language asserting that large AI spending is required for China competition also strengthened, rising by 13 points to 45. Anthropic accelerated this narrative by publicly accusing three Chinese AI companies, DeepSeek, MiniMax, and Moonshot, of setting up more than 24,000 fake accounts with Claude to improve their own models through distillation attacks. Anthropic is now arguing that these attacks undermine export controls, and that "without visibility into these attacks, the apparently rapid advancements made by these labs are incorrectly taken as evidence that export controls are ineffective." This reframes the China AI competition narrative from one about Chinese innovation to one about intellectual property extraction.

Our semantic signature tracking language asserting that the country which builds the best energy infrastructure will determine AI leadership declined by 61 points, the single largest weekly drop, from 279 to 218. While still well above average, this sharp pullback suggests that media attention is pivoting from energy capacity as the key determinant of AI dominance toward IP security, chip access, and deployment models. The Trump administration's planned "Tech Corps" initiative, which would deploy as many as 5,000 American volunteers and advisers to Peace Corps partner nations to steer developing countries toward U.S. AI hardware and away from Chinese technology, underscores how this competition is broadening well beyond models and chips.

Bubble Fears Persist as Dot-Com Comparisons Fade and Disruption Anxiety Spreads

The capital fueling this global competition is simultaneously stoking concern about financial sustainability. Perscient's semantic signature tracking language predicting that an AI investment collapse will crash overall markets rose by 7 points to 176, now 176% above the long-term mean and one of the most elevated readings in the dataset. Our signature tracking language asserting that businesses increasingly doubt large AI spending held flat at 113, also well above average. These readings reflect a persistent media narrative of investment overreach, but the specific framing is evolving in important ways.

ClearBlue Wealth's weekly commentary described how anxiety over a potential AI bubble fueled by relentless capital spending on data center infrastructure has "transitioned into a broader set of worries about industry-level disruption driven by rapidly advancing AI platforms." This is confirmed in the signature data: our semantic signature tracking language comparing AI spending to telecom overbuilding in the 1990s fell by 19 points to -7, moving from above average to below in a single week. The dot-com analogy appears to be losing resonance, a shift visible in social commentary too, where one observer noted the whiplash from "scaling is dead" to "DeepSeek means compute glut" to "all software and jobs will be disrupted tomorrow."

What has replaced the dot-com framing is more concrete. Last month, Anthropic's legal plugin release ignited an $830 billion global selloff in software and services stocks over six trading days. The sector has now lost around $2 trillion of market capitalization from its peak. The Financial Times reported that Big Tech's spending spree is reigniting bubble fears, noting that Alphabet, Microsoft, Amazon, Meta, and Oracle are expected to spend over $600 billion on AI-related development in 2026. New derivatives instruments have emerged as hedging tools, and the ECB has stepped up scrutiny of European banks' exposure to the AI sector.

Our signature tracking language asserting that massive AI infrastructure projects face doubts sits at 86, up by 1, while the signature tracking language predicting continued expansion of hyperscale infrastructure stands at 83, down by 3. These two readings are running nearly in parallel, reflecting a media environment that simultaneously acknowledges accelerating spending and questions its sustainability.

Our semantic signature tracking language asserting that promised AI efficiency improvements haven't occurred remains elevated at 47, though it moderated by 5 points this week. A National Bureau of Economic Research study published in February found that despite 90% of firms reporting no impact of AI on workplace productivity, executives still projected AI to increase productivity by 1.4% and output by 0.8%. The signature tracking language claiming that excessive AI rules will harm U.S. technological leadership fell by 27 points to -38, one of the sharpest weekly declines, likely reflecting how the current administration's deregulatory posture has reduced the perceived salience of that concern.

Fundstrat's Hardika Singh suggested that the software selloff "has gotten ahead of itself" and could end up like a "DeepSeek moment, where we forget about it a year later," while warning that "if they cannot revamp themselves, I think this is a rupture of the AI trade." The net picture is one of elevated but evolving anxiety: bubble language is intensifying while dot-com comparison language is declining, suggesting that the media narrative is maturing from a generic historical analogy to a more specific story about AI platforms actively threatening incumbent software and services businesses.

A Broadening Surge in White-Collar Job Disruption Language, Offset by Growing Discussion of New Job Creation

The threat to incumbent businesses carries direct implications for their workforces. Perscient's semantic signature tracking language predicting that AI will eliminate consulting industry positions stands at 279, the second-highest absolute value among all tracked signatures, after rising by 11 points. The signature tracking language predicting that AI will eliminate financial analysis roles rose by 27 points to 89, and the equivalent for data analysis positions rose by 26 points to 54, both among the week's largest increases. Our signatures for software job destruction (47, up by 7) and legal job destruction (45, flat) round out a picture in which all five white-collar disruption signatures sit above average and four of five strengthened this week.

These are not abstract fears. The $830 billion selloff triggered by Anthropic's legal plugin is a concrete market event connecting AI product launches directly to job-displacement anxiety. IBM became the latest casualty after Anthropic launched Claude Code Security, and the stock suffered its largest single-day loss since the dot-com era. CNBC reported that CrowdStrike, Palo Alto Networks, and Cloudflare also slumped amid questions about the durability of cybersecurity business models. An Anthropic executive told Business Insider that the recent selloff reflects how fast AI has been advancing, saying that the industry is "grappling with exponentials."

The consulting-specific reading of 279 is particularly striking. Anthropic's latest updates have stoked fears that agentic AI could threaten an industry reliant on selling enterprise software packages and advisory services. A Harvard Business School study published February 20 found that roles with potential for AI augmentation handle tasks that can be automated alongside tasks requiring human involvement, citing financial analysts and clinical neuropsychologists among those with high augmentation potential. A Citigroup report found that 54% of financial jobs have a high potential for automation, more than any other sector. OpenAI CEO Sam Altman told Fortune that "AI superintelligence at some point on its development curve would be capable of doing a better job being the CEO of a major company than any executive, certainly me."

Yet the narrative is not monolithically bleak. Our semantic signature tracking language asserting that AI will generate novel employment categories rose by 24 points to 34, one of the week's steepest climbs, while the signature tracking language claiming that AI will fundamentally change professional office jobs ticked up by 6 points to 5. The parallel rise of both job-destruction and job-creation language suggests that the media framing is evolving beyond simple displacement. As one social commentator noted, "AI now gives everyone the ability to compete with the big corp. Money is no longer a limiting factor in making things." IBM itself reversed course, tripling entry-level hiring after discovering that cutting junior talent pipelines created a different problem: no mid-level talent in three to five years.

Our signature tracking language asserting that autonomous agent capabilities advance language models rose by 14 points to 15, and the signature for multi-modal architecture improvement sits at 79, up by 5. Deloitte's 2026 insights classify agentic AI as a key trend: systems that plan and execute multi-step processes, going beyond chatbots by taking actions without constant human supervision. **Yet while nearly two-thirds of organizations are experimenting with AI agents, fewer than one in four have successfully scaled them to production, a gap that may explain why disruption signatures for clerical work (-23) and creative jobs (-5) remain muted.** The narrative is currently concentrated in high-skill professional services where agentic demonstrations have been most visible.

Despite the intensity of this language, the New York Times reported that code generators still require extensive oversight, and Fortune noted that experts agree that AI-related layoffs have been "insignificant, so far," with recent banking headcount reductions attributed to pandemic-era overhiring rather than automation. This disconnect between narrative and on-the-ground reality represents a key tension for the weeks ahead.