China poised to lift travel ban on Manus founders

Share
China poised to lift travel ban on Manus founders

Beijing is unwinding the last restraints of the AI industry's most consequential forced deal reversal, while fresh capital flows into the businesses that test models and wire the data centers running them.

China plans to lift its travel ban on Manus' co-founders, and CEO Xiao Hong intends to return to Singapore as the company nears the end of its separation from Meta, the Financial Times reports. The ban dates to April, when Beijing's NDRC ordered Meta to unwind its $2 billion-plus acquisition of the Singapore-incorporated agent startup — the first use of China's foreign-investment security review to reverse a completed cross-border AI deal — and barred the founders from leaving the country during the probe. We covered Manus' announcement that it would return to independent operation last week — Manus returns to independent operation after Meta unwind — which came with an unusual cost: user data generated after Meta closed the deal in December gets wiped on August 23–24 in affected jurisdictions.

The lift, when it comes, closes the circle on a saga that has defined the new rules of AI nationalism. Xiao Hong returning to Singapore restores the founding team's mobility just as Manus prepares to relaunch as a standalone player, with Tencent reportedly in talks to become its largest external shareholder at the same $2 billion valuation and revenue said to be approaching $500 million a year. The message for the industry is that China will force a deal undone and then let the company walk away — the founders' punishment ends when the transaction does.


Vals, the San Francisco startup that tests AI models on real-world professional work, raised a $40 million Series A led by a16z at a $400 million valuation, closing Wednesday. Existing backers 8VC, Pear VC, and Bloomberg Beta returned, joined by HRT Ventures and Next Ladder Ventures. Vals builds evaluations around actual jobs — finance analysts, lawyers, software engineers — with private test sets, limited runs to prevent contamination, and automated grading against expert standards; its own finance benchmark finds frontier models fail roughly 52 percent of real analyst tasks. The round came with three new products and, per a16z, eightfold revenue growth over 2025, a doubled customer count, and a tripled team in six months. The bet is that as models move from answering questions to doing work, the company that grades the work becomes the trust layer — a16z's partner describing Vals as building "the ratings agency for AI."


Point2 Technology closed a $136 million Series B for its radio-frequency data center interconnect tech, with Arm, LB Investment, and Maverick Silicon among the backers — joined across the round by Nvidia's NVentures, Bosch Ventures, UMC, and Molex. Point2's e-Tube platform pushes data over plastic waveguides using RF signals instead of copper or lasers, claiming about 10 times the reach of copper at comparable cost, roughly three times lower power than optics, and far lower latency. Interconnect has become a defining constraint of AI buildouts — copper dies at rack scale and optics are expensive to cool and power — and the round is the latest sign that investors, including Nvidia itself, are placing bets on a third path. Techmeme flagged the round as one of the week's biggest, and Crunchbase lists it among the top funding rounds of the period.

What to watch: whether Tencent's reported buy-in closes and how Manus positions itself as an independent agent player without Meta's distribution.

China's first forced unwind of an AI acquisition is nearly finished — does the travel-ban lift reset the rules for cross-border AI deals? Tell us in the comments.

Read more

Akhetonics says its all-optical CPU reaches a customer in 2026

Akhetonics says its all-optical CPU reaches a customer in 2026

Light-based computing keeps promising more than it delivers — but one Munich startup has just put a date on its bet, and the interview laying it out is doing the rounds on Hacker News this week. Akhetonics says it will deploy its first commercial machine with a major customer by the end of 2026, with several more planned for 2027. The company, founded by Michael Kissner and Leonardo Del Bino, is building a computer where data enters as light, is switched as light, and circulates through memory

The Week in AI — October 5–11, 2026

The Week in AI — October 5–11, 2026

Every big claim this week turned out to rest on fine print more interesting than the headline: revenue only the company reporting it can define, safety tests sandboxed while the product keeps the web, and a Pentagon phase-out nobody would confirm until reporters kept asking. The week's top 5 1. OpenAI's revenue was $20 billion below the numbers everyone quoted — and the gap was definitional. The Financial Times reported Thursday that OpenAI's annualized revenue runs roughly $20 billion unde

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Russia's largest tech company is learning what the AI era's infrastructure war looks like from the receiving end — three data centers in four days, and with them much of the cloud layer Russian businesses run on. A Ukrainian drone strike knocked out Yandex's data center in Vladimir early Sunday morning, the third of the company's facilities hit since October 8. The site — reported at roughly 50 MW and designed for about 2,880 server racks — stopped operating completely after the attack, Yandex

Agent teams cost up to 5x more, barely score higher

Agent teams cost up to 5x more, barely score higher

The multi-agent hype train hit a benchmark this weekend — and the grid and the trucking regulators had quiet weeks of their own. Vals AI put agent teams head-to-head with single agents on its Vibe Code Bench, and the teams cost between 1.8 and 5.1 times more for almost no extra quality. The evals company ran GPT-6 Sol and Claude Opus 5.5 solo and in teams across 50 apps at two reasoning efforts; out of four comparisons, only one was statistically significant — Sol at medium effort, where the t