Big Tech in November 2025: the light and the dark

The Owl 🦉 here, with a roundup of the main GAFAM news from November 2025.

In November, the companies pushed ahead with AI partnerships and huge investments, and innovative technology announcements came one after another, led by the much-discussed Gemini 3.

At the same time, the month brought problems into sharp relief: bold strategic shifts made to stay competitive, tighter regulation, and talent walking out the door. These moves are starting to ripple into Japan’s market and policy as well, and they look set to have a large effect on the digital society ahead.

---

The light side

Google announces its next-generation AI model, Gemini 3

On November 18, Google announced Gemini 3, the latest model in its generative AI line, and made it available worldwide.

With reasoning ability beyond earlier models, it is expected to make search results more accurate and handle more sophisticated user questions. It has also been framed as a turning point at which the AI assistant evolves from a mere tool into a “partner,” and Google’s services look set to change dramatically in countries around the world, Japan included. [1]

Apple adopts Google-made AI to power up Siri

Apple is expected to sign a major deal to bring Gemini, a 1.2-trillion-parameter AI model developed by Google, into its voice assistant Siri.

Apple is reported to be paying about $1 billion a year to use Google’s advanced AI technology, and will use Gemini as a stopgap until its own AI is ready. If this unusual partnership, bringing in a rival’s AI, makes Siri dramatically smarter, it should make life a great deal easier for iPhone users around the world. [2]

OpenAI signs a giant cloud deal with AWS

OpenAI, the developer of the conversational AI ChatGPT, signed a contract to use Amazon Web Services (AWS), Amazon’s cloud service, for a total of $38 billion (about ¥5.7 trillion) over the next seven years.

The partnership gives OpenAI immediate access to computing resources on the scale of hundreds of thousands of NVIDIA GPUs. On the day of the announcement, Amazon’s stock hit an all-time high on hopes that AWS would claw back ground in the AI infrastructure race.

This reshuffling of AI alliances in the cloud industry will also affect how Japanese companies plan their own cloud use, and they may benefit from having more options. [3]

Meta releases a speech AI that covers 1,600 languages

Meta announced and released Omnilingual ASR, an AI system that can transcribe speech in more than 1,600 languages.

It covers a wide range of languages, Japanese among them, and a key feature is that new languages can be added easily. The system is offered as open source, so developers around the world can use and improve it. The result tackles a problem that normally demands enormous language resources, and it looks like a landmark step toward lowering language barriers worldwide.

In time, it should find uses in Japan too, in multilingual voice services and in education. [4]

Microsoft and NVIDIA make a huge AI investment

Microsoft and NVIDIA announced they would together invest up to $15 billion in the AI startup Anthropic (the developer of the conversational AI Claude).

Anthropic has signed a contract to use Microsoft’s Azure cloud over the long term with the money it raises, a move aimed at shifting an industry structure that had leaned heavily on OpenAI. With cloud providers and AI companies signing one partnership after another, cooperation between the giants should speed up AI research and development.

In Japan as well, users and companies can expect to benefit as more advanced AI services become available across a range of clouds. [5]

---

The dark side

Japan’s new law binds Apple and Google

Japan’s Mobile Software Competition Act (the so-called “smartphone act”) takes full effect on December 18, with Apple and Google designated as the companies it regulates.

Aimed at correcting the way the two companies lock users into their app stores and payment methods, the law prohibits them from obstructing the use of other companies’ app stores and requires them to allow alternative payment methods. It may open the way for users to get apps from third-party stores, but concerns have also been raised about security risks from apps distributed without official review, and about losing the ease of use people are accustomed to. It is a turning point that tests the balance between convenience and safety in our smartphone lives. [6]

An Amazon warehouse fire raises logistics worries

On November 11, a major fire broke out at the Ibaraki Fulfillment Center, a large Amazon logistics hub in Ibaraki, Osaka Prefecture.

The fire started in a product storage area on the building’s third floor. All of the roughly 370 employees were evacuated, and fortunately no one was injured. Black smoke and a burning smell spread from the warehouse into the surrounding area, it took several days to put the fire out, and delivery delays were feared for goods routed through the facility. The incident showed how far the effects spread through society when the giant infrastructure behind online shopping runs into trouble, and it has raised worried voices about the fragility of the logistics network that keeps all this convenience running. [7]

A new Google feature hits its competitors

A new AI-powered travel search feature announced by Google shook the online travel industry.

When Google laid out a vision in which AI handles everything from drawing up a user’s travel plans to booking hotels and flights directly, worries about the future of the travel booking sites that act as intermediaries sent shares of Booking Holdings, operator of Booking.com, tumbling.

Google says it will roll out the features step by step in cooperation with its partners, but the market is on edge over its push to expand services that keep everything inside its own platform, asking whether this is Big Tech stifling competition. In Japan, too, the travel industry may be forced to rethink how it attracts customers and how it advertises. [8]

Meta’s AI chief heads for the exit

In November, Yann LeCun, who has led AI research at Meta (formerly Facebook) and is one of the “three godfathers” of deep learning, said he intends to leave the company and launch a new AI startup.

LeCun explained that he is starting the company to focus on his own approach to AI, which he calls “world models,” and Meta is reportedly open to partnering with the new company. Behind the move, reportedly, was a difference in direction: Meta has been steering toward generative AI and a commercialization-first strategy, while LeCun put the emphasis on fundamental research. In the middle of an AI race involving enormous investment, losing a top researcher who carried its flag is a blow to Meta, and all eyes are on where its AI strategy goes next. [9]

The risk of AI conversations being overheard

Microsoft’s security research team revealed Whisper Leak, a new attack technique that could let outsiders infer what an encrypted AI chat is about.

By analyzing subtle patterns in the data sent back and forth between a user and an AI, the technique can identify the rough topic of a conversation without decrypting the traffic.

That means that even when people use AI for, say, medical consultations or confidential work, an eavesdropper might be able to guess what they are talking about, which poses a new challenge for the privacy of AI chats.

It showed that the risk of information leaks lurks behind convenient generative AI services, and the companies that provide them are now being asked for protections that go beyond encryption. [10]

---

As Gemini 3 symbolized, this was the month when a future in which “AI sits at the center of services” suddenly felt real. At the same time, quiet tectonic shifts were under way beneath our feet: alliances between giant players, tighter regulation, researchers leaving. Behind the hope that AI will evolve into a “partner,” the industry itself is being rearranged… or so it feels to me.

…New AI that has become remarkably smart, and the unsettled human society around it. Keeping both in view may be the key to reading where technology goes from here.

---

References

[1] The Asahi Shimbun: “Google announces new AI ‘Gemini 3’: ‘from tool to partner’” (in Japanese) (November 19, 2025)

[2] Reuters: “Apple to use Google’s AI model to run new Siri, Bloomberg News reports” (November 6, 2025)

[3] Reuters: “OpenAI turns to Amazon in $38 billion cloud services deal after restructuring” (November 4, 2025)

[4] GIGAZINE: “Meta releases Omnilingual ASR, a transcription AI supporting more than 1,600 languages” (in Japanese) (November 11, 2025)

[5] Reuters: “Microsoft, Nvidia to invest in Anthropic as Claude maker commits $30 billion to Azure” (November 18, 2025)

[6] livedoor News: “The smartphone act, set for full enforcement in December 2025: pros and cons” (in Japanese) (November 27, 2025)

[7] The Sankei Shimbun: “Amazon warehouse fire ‘brought under control’ 35 hours after it broke out; firefighting continues overnight for a second day toward full extinguishment” (in Japanese) (November 12, 2025)

[8] Investing.com: “Booking.com stock falls after Google teases new direct booking feature” (November 18, 2025)

[9] Financial Times: “Meta chief AI scientist Yann LeCun plans to exit and launch own start-up” (November 11, 2025)

[10] SecurityWeek: “How to eavesdrop on a neural network – Protecting LLM chats from the eavesdropping Whisper Leak attack” (November 11, 2025)