This question has a cousin concerning nationalization of labs outside of North America, EU27, UK or Australia, by end of 2021.
The AI lab can be a for-profit/financial subsidiary of a for-profit (such as DeepMind) or non-profit (such as OpenAI), but academic labs do not count. The lab must have had >=5 papers at major AI conferences (this excludes situations like the question resolving positive due to General Motors being nationalised and having an applied AI team).
The take-over must not involve compensation to shareholders, close to an estimated fair market price, for the loss of the net worth of seized assets and potential income.
The date of nationalization will be taken to be the date that a credible government outlet announces the intention to nationalize the lab, OR the date when another credible news outlet reports the nationalization, and the question will retroactively close one month prior to this date.
The location of the lab will be determined by the nationalizing country (even if the lab has subsidiaries, offices, or headquarters in another country).
Treatment of other special cases:
Resolves positive if a government takes operations control of an AI lab, but not ownership, such that the stockholders still get the profits, but can't make decisions (see this for a historical US example).
Resolves positive if the parent company of an AI lab is nationalized for non-AI reasons, e.g. US nationalizes Facebook because of privacy concerns.
The quasi-nationalization seen in China is not sufficient to resolve positive:
Communist Party committees have been installed at many tech firms, reviewing everything from operations to compliance with national goals. Regulators have been discussing taking a 1 percent stake in some giants, including Alibaba and Tencent, along with a board seat. Tech companies have been widely encouraged to invest in state-owned firms, in the hopes of making them more productive.