Taiwan's president Lai Ching-te during a visit to the Songshan military airbase in Taipei on March 21, 2025. (Photo: I-Hwa Cheng/AFP/Getty Images)

Taiwan has put two major Chinese tech companies on a trade blacklist.

On June 10, Shenzhen-based Huawei and Shanghai’s SMIC—two major players in China’s AI chip war—and some of their subsidiaries were added to a “strategic high-tech commodities entity list” that includes hundreds of firms based in Russia, Pakistan, Iran, Myanmar, and yes, mainland China.

Taiwanese companies must now seek approval from the island’s government before they can ship anything to the companies on the list.

The administration told Bloomberg that the move is intended “to combat arms proliferation and address other national security concerns.”

Translation: We don’t want listed companies to access Taiwanese chipmaking tech and facilities.

There’s reason for concern. Two years ago, several Taiwanese companies were reported to have been helping Huawei build chip plants on mainland China.

Meanwhile the largest contract chip manufacturer—TSMC, partner to Apple, Nvidia, AMD, Qualcomm, and others—remains headquartered in Taiwan.

Cross-strait relations between Taiwan and China, never great, have deteriorated this year in light of the broader geopolitical landscape. 

In March, Taiwan’s leader Lai Ching-te labeled China a “foreign hostile force,” to which China replied: “Those who play with fire will surely be burned.” —AN

Amazon is taking its AI investments down under.

The American megaretailer plans to invest AU$20 billion (or about US$13 billion) in Australia to develop data center infrastructure for AI and cloud computing.

The investment—the largest ever in Australia by a major tech company, according to PM Anthony Albanese—would begin this year and run through 2029.

“This is the largest investment our country has seen from a global technology provider, and is an exciting opportunity for Australia to build AI capability using secure, resilient infrastructure,” Albanese said, adding that the funds would “support complex AI and supercomputing applications.”

The funds would be directed at Amazon’s data center facilities in Sydney and Melbourne.

Amazon would also invest in three new solar farms in Victoria and Queensland, operated by European Energy, to supply more than 170 megawatts of power to its operations.

Amazon said earlier this year that it planned to spend $100 billion on capital expenditures in 2025, most of it earmarked for AI capabilities within its Amazon Web Services division. —AN

OK, OK: So tech deals aren’t the hottest thing around right now, given the global economic landscape.

But zoom into AI—and specifically data infrastructure—and you’ll find that the M&A flames burn bright.

A recent Reuters report notes that the companies who process data used to build AI models “have become highly sought after” by Big Tech companies like Meta, Salesforce, and ServiceNow who seek to compete with AI model category leaders OpenAI, Anthropic, and Google.

According to markets platform Dealogic, tech deals accounted for about 25% of global M&A activity in the first five months of the year—$421 billion of $1.67 trillion, to be exact. 

That’s up from a 20% share last year and 17% share the year before that.

But here’s the kicker: Almost three quarters of those deals, in terms of value, were AI software—particularly enterprise AI software, and especially those focused on data analytics and infrastructure.

Some of these companies you’ve heard of (because this newsletter has told you), some you haven’t: Boomi, Collibra, Confluent, Dataiku, Fivetran, Matillon, Qlik, Signa Computing. 

But a firm is not yet a household name for you, rest assured, it is for an acquisitive tech CEO who needs an AI edge.

As Dataiku chief Florian Douetteau put it at Fortune’s Brainstorm AI gathering in London last month: “Many CEOs realize that they have to do things about AI and are almost, to some extent, fearful of losing their job or their position because of AI, because of not doing enough.” —AN

—Anne Wojcicki wins back 23andMe. The former CEO’s $305 million bid outmaneuvers Regeneron.

—More BT Group cuts on the horizon? AI advancements could make them a reality, the telco’s CEO says.

—Meta’s Llama 3.1 can recall large parts of books. AI content memorization: not so unusual after all.

—New York state passes frontier AI bill. The safety and transparency requirements await Gov. Hochul's signature.

—Are LLMs really ready for the doctor’s office? Chatbots are book-smart but not so good with real human patients, a new study says.

—All hail the influencer. When ad budgets get tight, more capital flows to so-called creators.

—U.S. SEC rescinds rules. Fourteen of ‘em, including those about crypto, proposed during the Biden administration.

—Salt Lake City mulls using AI for 9-1-1 calls. Could handle up to 30% of the non-emergency volume.