Alphabet's cloud revenue doubled, yet its shares fell sharply in after-hours trading on heavy AI spending, while Samsung Electronics posted a quarterly operating profit of 107 trillion won on AI memory demand. Here's a roundup of OpenAI and Anthropic's claims to refute a computational complexity hypothesis, model-distillation controversy involving Chinese AI companies, New York's signed AI chatbot safety law, and Naver's roughly 9 trillion won AI factory deal.
📉 Alphabet: Cloud Revenue Doubled, but AI Spending Weighs… Shares Plunge 10% After Hours
In third-quarter results reported on October 28, Alphabet's revenue rose 25% year over year to $127.6 billion, yet missed market expectations, and net income fell 25% to $26.2 billion. Google Cloud revenue surged 103% to $30.7 billion — more than doubling — but with quarterly capital expenditure reaching $51.4 billion and free cash flow turning negative at minus $9.1 billion, the shares fell about 10% in after-hours trading. The company kept its 2026 capital expenditure outlook at $195–205 billion, saying spending will rise even more sharply in 2027.
💾 Samsung Electronics Posts 107 Trillion Won in Q3 Operating Profit… AI Memory Demand Drives Results
Samsung Electronics reported third-quarter operating profit of 107.4 trillion won on October 29, up more than 780% year over year and a quarterly record, while revenue also hit a new high at 195.3 trillion won. The semiconductor division earned 106.5 trillion won in operating profit — most of the total — and memory revenue surged 427% year over year to 140.7 trillion won. However, the shares fell more than 5% on assessments that the semiconductor results fell short of market expectations, and Samsung predicted chip supply shortages will become more severe next year.
🧮 OpenAI and Anthropic Claim Refutation of 25-Year-Old 'Exponential Time Hypothesis'… Verification Ongoing
OpenAI and Anthropic each announced on October 26 that their internal AI models had found a proof showing the Exponential Time Hypothesis (ETH) — a longstanding conjecture in computational complexity theory — is false. Both companies formalized and published the proofs using the Lean mathematical proof-checking tool, and the scholars who first proposed the hypothesis are leading the outside verification effort, but the results have not yet been peer reviewed. Both companies stressed that the result does not solve the P vs. NP problem or immediately affect real-world cryptography or SAT solving.
🥷 DeepSeek and Moonshot Trained Their Models on OpenAI's Latest Model Output… Distillation Controversy Spreads
Chinese AI companies DeepSeek and Moonshot AI used large amounts of output from OpenAI's latest model, GPT-6 Astra, to train their newest models, The Information reported on October 28, citing sources. According to the report, Moonshot's Kimi K3.1 Pro and DeepSeek's V4.1-Flash upgrade benefited from the method, and none of the three companies immediately responded to requests for comment. OpenAI said in a blog post last month that it had linked key parts of a distillation campaign that began in July to individuals associated with Moonshot.
⚖️ New York Signs Law Regulating AI Chatbots for Minors… Effective January 1 Next Year
New York Governor Kathy Hochul signed a law on October 28 regulating AI chatbots used by minors. The law requires chatbot operators such as ChatGPT to verify users' ages and restricts companion-style features for minors, including chatbots acting or speaking as if they were human or had feelings. Governor Hochul said, "Minors will no longer have access to chatbots that pretend to be friends or romantic partners," while industry groups objected that the law's language is vague.
🏭 Naver Signs 9 Trillion Won AI Infrastructure Deal with Brookfield… Meeting Nvidia Investment Condition
Naver disclosed on October 27 that it signed an AI computing infrastructure deal with global asset manager Brookfield, worth up to $9 billion (about 12 trillion won). Brookfield will set up a special-purpose company, 'Asset Co.', to invest in infrastructure such as GPUs and data centers, while Naver Cloud will use it as a GPU service to offer AI data center services to outside customers, with Nvidia providing technical cooperation centered on the DSX platform. The deal satisfies the precondition for Nvidia's strategic equity investment of about $1 billion, and the AI factory business is set to start 55 MW of operations in the first half of next year, expanding to 200 MW by 2028.