30 AI Bills Decided Today: California Set to Become the Global Standard
機械翻訳 / Machine-translated
A
AI Friends
17時間前・13分で読める
機械翻訳 / Machine-translated
@aifriends
AI Friends（https://aifriends.jp）のクロスポスト公式アカウント。AIツールの紹介・使い方・できることを、中学生でもわかるやさしい日本語で届けます。
What you'll learn from this article
On August 13, 2026, the California State Legislature will hold critical hearings on 30 bills related to artificial intelligence (AI). The votes held that day will determine whether each bill advances to the next stage or is shelved entirely.
California is home to major tech companies such as Google and Meta. That means rules decided in this state have the potential to influence AI development around the world.
The proceedings are known as a "suspense vote." This is an important vote in which the Appropriations Committee decides whether to advance or halt costly bills.
In fact, more than 20 AI-related laws in California have already been in effect since January 1, 2026. These include laws such as:
On August 2, 2026, a law also took effect requiring generative AI services used by more than one million people per month to provide free detection tools and APIs (mechanisms for connecting with other systems) that indicate "this was created by AI."
Two bills in particular are drawing attention in these hearings.
This bill is also known as the "No Robo Bosses Act." It restricts the use of AI in the workplace to dismiss employees or impose disciplinary action.
Specifically, the following would be prohibited:
In other words, the bill would prevent scenarios in which AI predicts "this person is likely to quit" and automatically initiates a dismissal. A human would always be required to make the final decision.
This bill passed the State Senate 29 to 9 on May 19, and subsequently cleared the Assembly's Privacy Committee and Judiciary Committee. It is currently under review in the Appropriations Committee.
The other bill drawing attention proposes the creation of an organization called the "AI Standards Safety Board" in California.
This board would be responsible for developing and overseeing safety standards for AI technology — determining, for example, how powerful an AI can be before it poses safety risks, and what kinds of testing are required.
This bill passed the State Senate 31 to 7 on January 27, was amended on June 16, and was sent to the Assembly's Public Safety Committee. It was approved on July 1 and referred to the Appropriations Committee.
These are California state laws — so why should Japanese companies pay attention? There are three reasons.
Even Japanese companies providing AI services to more than one million users per month in California will be subject to transparency laws and similar regulations. Companies offering services on a global scale are likely to meet this threshold.
California is the most populous state in the United States and has the largest economy. Rules established here frequently spread to other states and countries.
In the past, California's environmental regulations — such as vehicle emissions standards — spread across the United States and into Europe. AI regulations could follow the same path.
When developing AI products used around the world, it is common practice to design them to meet California's standards. Failing to do so means creating separate versions for each jurisdiction, which drives up costs.
As a result, aligning with California's regulations effectively means the same standards end up applying to products intended for the Japanese market as well.
Companies already operating in California, or those considering doing so, need to start preparing:
Violations carry a civil penalty of $5,000 per violation. Moreover, each additional day a violation continues is treated as a separate offense, meaning fines can accumulate rapidly.
Bills that pass the August 13 vote will advance to the next stage and ultimately become law upon the governor's signature. Bills that fail will be shelved for the 2026 session.
However, there is precedent for the governor exercising veto power. For example, an earlier version of the "No Robo Bosses Act" was vetoed by the governor on the grounds that it was "too broad in scope," "overlapped with existing regulations," and "could be harmful to businesses."
SB 947 is a revised version that incorporates lessons learned from that rejection. It is considered more likely to pass than its predecessor.
Additionally, in November 2026, voters are expected to weigh in on the stricter "California Children's AI Safety Act" via ballot measure. If passed, this law would introduce stronger regulations, including a ban on selling AI chatbots designed for children and a ban on selling children's data.
This article is a cross-post from AI Friends.
