Cheap AI could derail OpenAI and Anthropic's IPOs

Cnbc · 🇨🇳 public China · May 20, 2026 · 1178 words · By Deirdre Bosa; Jasmine Wu

Enterprise AI Adoption AI Market Economics US-China Tech Competition

headphones Listen to the eFinder podcast briefing

Generate a natural audio summary of this story

Daily briefing

What to know about Enterprise AI Adoption

This earnings season, the cost of AI started showing up in the numbers.

Claims checked 17

Techniques found 2

Topics 3

Coverage spectrum

Coverage gap: Low Left coverage

Left0%

Center100%

Right0%

6 sources compared across this story cluster. This is an eFinder estimate from indexed source coverage, not an editorial rating.

What happened

This earnings season, the cost of AI started showing up in the numbers.

Why it matters

Meta, Shopify, Spotify, and Pinterest all flagged rising AI and inference costs as a drag on margins.

Common ground

Shopify said economies of scale were "partially offset by increased LLM costs." This is the bill coming due for the pricing model that underpins OpenAI's and Anthropic's expected IPO valuations, both projected north of $800 billion.

Perspective signals

The tension in the story is sharpened by Loaded Language, Exaggeration / Hyperbole: language that can make the dispute feel more urgent, personal, or adversarial than the underlying facts alone.

Follow-up questions

What new context would change how readers understand this Enterprise AI Adoption story?
What evidence would most clearly confirm or weaken the claim that Models from other Chinese labs, including Moonshot, Xiaomi, and Zhipu, have shipped at similar capability levels in the past four months?
How does this story connect Enterprise AI Adoption with AI Market Economics over the next few days?

open_in_new Read the original article: https://www.cnbc.com/2026/05/20/cheap-ai-could-derail-openai-and-anthropics-ipos…

psychologyPropaganda Techniques Detected

eFinder identified 2 propaganda techniques in this article. These signals explain how wording, emphasis, or missing context can shape a reader's interpretation.

warning

Loaded Language 80% confidence

Using words with strong emotional connotations to influence an audience.

Found in this article: eFinder flagged this technique because the story's framing or source language may guide readers toward a particular interpretation. Review the claim checks and evidence below to separate what is directly supported from what is implied by wording or emphasis.

Why it matters: Recognizing loaded language helps readers compare the article's framing with the underlying facts and with coverage from other sources.

warning

Exaggeration / Hyperbole 70% confidence

Overstating facts or claims to create a stronger emotional response.

Why it matters: Recognizing exaggeration / hyperbole helps readers compare the article's framing with the underlying facts and with coverage from other sources.

fact_checkClaims Checked

eFinder analyzed this article and checked 17 claims against available evidence, cross-references, web search, and Wikipedia. Here is what the fact-checking layer found.

schedule Pending 7

verified Verified By Reference 3

help Insufficient Evidence 2

info Single Source 2

report Misleading 1

cancel Disputed 1

check_circle Corroborated 1

help

Claim 1: “Models from other Chinese labs, including Moonshot, Xiaomi, and Zhipu, have shipped at similar capability levels in the past four months.”

INSUFFICIENT EVIDENCE

No evidence was provided for this claim.

report

Claim 2: “Some 45% of companies surveyed by cloud cost firm CloudZero said they spent more than $100,000 a month on AI in 2025, up from 20% the year before.”

MISLEADING

The evidence from CloudZero mentions that organizations planning to invest over $100,000 monthly 'more than doubled from 20%', but it does not explicitly state that 45% of companies spent more than $100,000 a month in 2025. The numbers provided in the claim are not directly supported by the snippet.

menu_book

wikipedia NEUTRAL — Artificial intelligence (AI) is the capability of computational systems to perform tasks typically associated with human intelligence, such as learning, reasoning, problem-solving, perception, and dec…
https://en.wikipedia.org/wiki/Artificial_intelligence

menu_book

wikipedia NEUTRAL — Cloud computing is defined by the International Organization for Standardization (ISO) as "a paradigm for enabling network access to a scalable and elastic pool of shareable physical or virtual resour…
https://en.wikipedia.org/wiki/Cloud_computing

menu_book

wikipedia NEUTRAL — Google Cloud is a suite of cloud computing services offered by Google that provides a series of modular cloud services including computing, data storage, data analytics, and machine learning, alongsid…
https://en.wikipedia.org/wiki/Google_Cloud_Platform

+ 3 more evidence sources

schedule

Claim 3: “Figma is selling features that cut customers' token consumption by 20 to 30%.”

PENDING

This claim was extracted as a checkable statement from the article. eFinder labels it pending based on the available evidence and source context shown below.

info

Claim 4: “OpenAI's confidential filing coming as soon as this week”

SINGLE SOURCE

The search results for OpenAI's filing are generic Wikipedia or GitHub pages and do not mention a confidential filing coming 'this week'.

travel_explore

web search NEUTRAL — OpenAI is an American artificial intelligence (AI) research organization headquartered in San Francisco, consisting of a for-profit public benefit corporation (PBC) partially controlled by a nonprofit…
https://en.m.wikipedia.org/wiki/Openai

travel_explore

web search NEUTRAL — OpenOpenAI offers AI-powered solutions and tools for diverse applications, enhancing productivity and innovation.
https://www.open-openai.com/app

travel_explore

web search NEUTRAL — OpenAI has 248 repositories available. Follow their code on GitHub.
https://github.com/openai/

verified

Claim 5: “Shopify said economies of scale were "partially offset by increased LLM costs."”

VERIFIED BY REFERENCE

The evidence provided for Shopify is generic marketing material or general guides on profit margins; there is no specific quote or report stating that LLM costs partially offset economies of scale.

menu_book

wikipedia NEUTRAL — Microsoft Copilot is a generative artificial intelligence chatbot developed by Microsoft AI, a division of Microsoft. Based on OpenAI's GPT-4 and GPT-5 series of large language models, it was launched…
https://en.wikipedia.org/wiki/Microsoft_Copilot

menu_book

wikipedia NEUTRAL — Ruby is a general-purpose programming language. Its design puts an emphasis on programming productivity and simplicity. In Ruby, everything is an object, including primitive data types. Yukihiro "Matz…
https://en.wikipedia.org/wiki/Ruby_(programming_language)

menu_book

wikipedia NEUTRAL — Toloka is a Dutch multinational data services company based in Amsterdam, Netherlands. It delivers human-in-the-loop annotation and evaluation work that supports the development of generative AI and l…
https://en.wikipedia.org/wiki/Toloka

+ 3 more evidence sources

cancel

Claim 6: “OpenAI's and Anthropic's expected IPO valuations, both projected north of $800 billion.”

DISPUTED

Sources provide conflicting valuation numbers. One source mentions a combined $2 trillion wave, while another specifically estimates Anthropic at $300-350bn and OpenAI at $500bn. Neither source confirms a projected IPO valuation 'north of $800 billion' for *each* company.

menu_book

wikipedia NEUTRAL — Cerebras Systems Inc., headquartered in Sunnyvale, California, develops semiconductors, supercomputers, and related software to power artificial intelligence deep-learning applications such as inferen…
https://en.wikipedia.org/wiki/Cerebras_Systems

menu_book

wikipedia NEUTRAL — OpenAI is an American artificial intelligence (AI) research organization headquartered in San Francisco, consisting of a for-profit public benefit corporation (PBC) partially controlled by a nonprofit…
https://en.wikipedia.org/wiki/OpenAI

menu_book

wikipedia NEUTRAL — The Six AI Tigers (Chinese: 大模型六小虎; pinyin: Dà móxíng liù xiǎohǔ; lit. 'Six Little Tigers of Large Models') is a collective designation for six China-based artificial intelligence startup companies es…
https://en.wikipedia.org/wiki/Six_AI_tigers

+ 3 more evidence sources

schedule

Claim 7: “On OpenRouter... Chinese models went from about 1% of usage in 2024 to more than 60% in May.”

PENDING

This claim was extracted as a checkable statement from the article. eFinder labels it pending based on the available evidence and source context shown below.

verified

Claim 8: “Meta, Shopify, Spotify, and Pinterest all flagged rising AI and inference costs as a drag on margins.”

VERIFIED BY REFERENCE

The provided evidence mentions general AI coding startup struggles and Shopify's general business, but does not contain specific reports from Meta, Shopify, Spotify, and Pinterest collectively flagging AI inference costs as a drag on margins.

menu_book

wikipedia NEUTRAL — Reddit ( RED-it, formerly stylized reddit) is an American proprietary social news aggregation and forum social media platform. Registered users (commonly referred to as "redditors") submit content to…
https://en.wikipedia.org/wiki/Reddit

menu_book

wikipedia NEUTRAL — An Internet entrepreneur is an owner, founder or manager of an Internet-based business. This list includes Internet company founders and people brought on to companies for their general business or ac…
https://en.wikipedia.org/wiki/List_of_Internet_entrepreneurs

menu_book

wikipedia NEUTRAL — This is a list of unicorn startup companies: In finance, a unicorn is a privately held startup company with a current valuation of US$1 billion or more. Notable lists of unicorn companies are maintain…
https://en.wikipedia.org/wiki/List_of_unicorn_startup_compan…

+ 3 more evidence sources

schedule

Claim 9: “Cohere's revenue grew sixfold last year selling into [banks, defense agencies, and other regulated industries].”

PENDING

This claim was extracted as a checkable statement from the article. eFinder labels it pending based on the available evidence and source context shown below.

schedule

Claim 10: “Reflection AI raised at a multibillion-dollar valuation specifically to build American open-source models for enterprises”

PENDING

This claim was extracted as a checkable statement from the article. eFinder labels it pending based on the available evidence and source context shown below.

schedule

Claim 11: “the U.S. government's AI Safety Institute... documented that downloads have risen nearly 1,000% since the R1 release in January 2025.”

PENDING

This claim was extracted as a checkable statement from the article. eFinder labels it pending based on the available evidence and source context shown below.

schedule

Claim 12: “every release of a new frontier model, including GPT-5.5 last month, has driven a surge in API and product usage”

PENDING

This claim was extracted as a checkable statement from the article. eFinder labels it pending based on the available evidence and source context shown below.

help

Claim 13: “DeepSeek... released a preview of its next-generation model last month that matches or nearly matches the latest from OpenAI, Anthropic, and Google on coding, agentic, and knowledge benchmarks.”

INSUFFICIENT EVIDENCE

No evidence was provided for this claim.

info

Claim 14: “At its I/O developer conference this week, CEO Sundar Pichai said "many companies are already blowing through their annual token budgets, and it's only May"”

SINGLE SOURCE

The evidence provides general information about Sundar Pichai and Google I/O, but does not contain the specific quote regarding companies 'blowing through their annual token budgets'.

travel_explore

web search NEUTRAL — Pichai Sundararajan, better known as Sundar Pichai, is an American business executive who has been the CEO of Google since 2015 and the CEO of its parent company Alphabet Inc. since 2019.
https://en.wikipedia.org/wiki/Sundar_Pichai

travel_explore

web search NEUTRAL — Google & Alphabet CEO Sundar Pichai sits down exclusively with Bloomberg Originals Host & Executive Producer Emily Chang to discuss the future of search, reb...
https://www.youtube.com/watch?v=5puu3kN9l7c

travel_explore

web search NEUTRAL — Картинки. Войти. Google. Расширенный поиск.
https://www.google.com/

verified

Claim 15: “For each lab's most capable model: Anthropic's Claude came in at $4,811. OpenAI's ChatGPT: $3,357. DeepSeek: $1,071. Kimi: $948. Zhipu's GLM: $544.”

VERIFIED BY REFERENCE

While Artificial Analysis is identified as a benchmark platform, the specific cost figures ($4,811, $3,357, etc.) are not present in the provided evidence snippets.

menu_book

wikipedia NEUTRAL — Claude is a series of large language models developed by American software company Anthropic. Claude was released as a AI chatbot in March 2023. It is also used in AI-assisted software development. Cl…
https://en.wikipedia.org/wiki/Claude_(language_model)

menu_book

wikipedia NEUTRAL — Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd., doing business as DeepSeek, is a Chinese artificial intelligence (AI) company that develops large language models (LLMs).…
https://en.wikipedia.org/wiki/DeepSeek

menu_book

wikipedia NEUTRAL — A generative pre-trained transformer (GPT) is a type of large language model (LLM) that is widely used in generative artificial intelligence chatbots. GPTs are based on a deep learning architecture ca…
https://en.wikipedia.org/wiki/Generative_pre-trained_transfo…

+ 3 more evidence sources

check_circle

Claim 16: “If the largest Google Cloud customers shifted 80% of their workloads from frontier models to Gemini 3.5 Flash, Pichai said, they would save more than $1 billion a year.”

CORROBORATED

Two independent web search results explicitly confirm that Sundar Pichai stated shifting 80% of workloads to Gemini 3.5 Flash would save largest customers over $1 billion annually.

travel_explore

web search NEUTRAL — The first model from the new family, 3.5 Flash, is out now and already powering the Gemini app and AI Mode in Search.If they shifted 80 per cent of their workloads from other frontier models to 3.5 Fl…
https://www.theage.com.au/technology/google-unveils-new-sear…

travel_explore

web search NEUTRAL — If they shifted 80% of their workloads from other frontier models to 3.5 Flash, they’d save over $1 billion dollars annually. That is real savings they can pour back into their company.
https://blog.google/innovation-and-ai/sundar-pichai-io-2026/

travel_explore

web search NEUTRAL — Google shipped Gemini 3.5 Flash across AI Mode, Antigravity, Spark and enterprise products while asking developers to wait for Pro. The release puts speed under the agent layer first.
https://www.implicator.ai/gemini-3-5-flash-lands-inside-sear…

schedule

Claim 17: “In a policy paper released in May, [Anthropic] said U.S. models are only "several months ahead" of Chinese ones, and warned that Beijing is "winning in global adoption on cost."”

PENDING

This claim was extracted as a checkable statement from the article. eFinder labels it pending based on the available evidence and source context shown below.

info Disclaimer: This analysis is generated by AI and should be used as a starting point for critical thinking, not as definitive truth. Claims are verified against publicly available sources. Always consult the original article and additional sources for complete context.