Here are today's news for 19th of November about Microsoft, Anthropic, Innovation and Science. Let's dive into the details. Today in brief: On November 18, Microsoft unveiled seven AI models for coding, image editing, voice, and other tasks; On November 18, Anthropic priced its IPO at $83 a share, above its projected range, in a landmark public offering by a major AI model developer; A report published November 18 called for a statutory AI regulator with powers to pause high, risk training, block launches, and suspend services; On November 17, Alibaba was the only major technology company reported to have signed the UN, backed Antalya AI Pledge, which calls for voluntary steps to reduce AI’s environmental impact and On November 18, Nvidia forecast fourth, quarter revenue of about $130 billion, above analyst expectations, helping lift AI, related stocks amid concerns about demand. On November 18, Microsoft announced seven AI models at Ignite 2026, spanning coding, cybersecurity, image editing, speech, transcription and model routing. The lineup includes MAI Code 2 and Code 2 Flash, Cyber 2 Flash, Image 2.7, Realtime 1, Dragon Copilot Transcribe, built on MAI Transcribe 2, and Decision 1.1. Microsoft says Code 2 averaged 98% of GPT, 6.1 Sol’s performance across four coding benchmarks at about 40% of its price, and 96% of Claude Opus 5.5’s at 5% of the price. Cyber 2 Flash reportedly beat GPT, 6 Astra on CyberGym, E2E for under 2% of the cost per task; Realtime 1 starts at $0.01 per audio minute. Microsoft says Image 2.7 leads Artificial Analysis’s image, editing ranking at one, fifth the price of GPT Image 2.6 Sunburst. The launch sharpens Microsoft’s effort to reduce reliance on external AI providers and compete on enterprise cost: Code 2 is aimed at GitHub Copilot, while Dragon Copilot Transcribe already supports 10 million patient encounters monthly, with multilingual error rates halved. On November 18, 2026, Anthropic priced its IPO at $83 per share, above its marketed $70, $78 range, selling 980 million shares and raising about $81.3 billion. The company offered 956.4 million new shares; existing holders sold 23.6 million. At the offer price, Anthropic was valued at roughly $1.5 trillion, and its Nasdaq listing was scheduled for November 19 under ticker ANT. The deal’s scale, larger than SpaceX’s reported $75 billion offering, makes it a major test of public investors’ appetite for AI companies and the enormous costs of building them. Proceeds give Anthropic capital to fund computing infrastructure and model development, while the listing creates a public, market benchmark for rivals such as OpenAI and Google. Founded in 2021 by former OpenAI researchers, Anthropic developed the Claude family of AI models and expanded its business through paid chatbot access and enterprise services. On 18 November 2026, the Commons Business, Innovation, Science and Trade Committee published Guardrails for Growth: AI Safety and Security, identifying ten gaps in UK safeguards and urging an AI Safety and Security Bill in the next parliamentary session. It wants a statutory regulator, separate from the AI Security Institute (AISI), able to verify safety claims independently, pause high, risk training and testing, delay or prohibit launches, and suspend services when risks cannot be controlled. The report says no UK authority currently has power to pause frontier, model training or withdraw unsafe services; firms make those decisions themselves. That leaves OpenAI, Google DeepMind, Anthropic and Meta, all heard by the committee, facing potential new oversight, while clearer rules could reduce uncertainty for businesses adopting AI. The committee cites estimates that AI could make Britain’s economy up to £400 billion larger after ten years, and sees AI assurance as an export opportunity. On November 17, Reuters reported that Alibaba was the only major technology company to have signed the Antalya AI Pledge, launched November 16 at COP31 in Turkey. The voluntary initiative had 90 endorsements at launch, including nine countries, and calls for cleaner power, more efficient AI infrastructure, environmental, impact reporting and support for AI climate projects in developing countries. Its signatories face no new obligations; the International Telecommunication Union says it will remain open to COP32. The absence of major US firms leaves Alibaba unusually visible as AI’s energy and water demands draw scrutiny from governments, investors and communities near data centers. Google, Microsoft, Meta and Nvidia had not signed; Amazon pointed Reuters to its own targets. Alibaba’s existing figures give the pledge concrete context: in fiscal 2026, clean electricity supplied 73.6% of its self, built data, center power, while average power usage effectiveness reached 1.187. On November 17, Nvidia forecast fiscal fourth, quarter revenue of $130 billion, plus or minus 3%, for the quarter ending in January. The midpoint tops LSEG analysts’ $125.25 billion estimate; even the low end, $127.4 billion, is higher. The outlook followed fiscal third, quarter revenue of $115.84 billion, up more than 100% year over year, and adjusted earnings per share of $2.68. Nvidia expects fourth, quarter adjusted gross margin of 71%, 72%, near analysts’ 71.7% estimate. The forecast helped lift AI, linked shares on November 18, as investors looked for evidence that data, center spending can sustain the chip boom. Nvidia’s GPUs power major AI systems, so its orders signal demand for cloud providers and competitors alike. Yet its guidance assumes no data, center chip sales to China, where licensed Hopper sales contributed $2.4 billion in the quarter. Nvidia said demand elsewhere exceeds supply and raised its fiscal 2028 revenue, growth outlook to about 75%, from 70%. That's all for today's AI news briefing. Have a great day!