{"id":35559,"date":"2025-10-06T13:48:47","date_gmt":"2025-10-06T11:48:47","guid":{"rendered":"https:\/\/magazine.luxus-plus.com\/?p=35559"},"modified":"2025-10-06T13:48:47","modified_gmt":"2025-10-06T11:48:47","slug":"artificial-intelligence-ai-is-now-regulated-in-california","status":"publish","type":"post","link":"https:\/\/magazine.luxus-plus.com\/en\/artificial-intelligence-ai-is-now-regulated-in-california\/","title":{"rendered":"Artificial intelligence (AI) is now regulated in California"},"content":{"rendered":"

More than a year after European legislation regulating artificial intelligence came into force, the state of California, the birthplace of this technology, has also just enacted a similar law… While the Trump administration is hostile to any restrictions in the AI race, the initiators of the so-called \u201cSB 53\u201d law have sought to strike the right balance so as not to stifle innovation but to avoid its excesses.<\/p>\n

<\/p>\n

 <\/p>\n

California<\/b> puts a stop to the uncontrolled development of AI<\/b>.<\/p>\n

 <\/p>\n

As the birthplace of artificial intelligence, with the headquarters of tech leaders such as Google, Meta, OpenAI (ChatGPT) and Anthropic<\/b> (founded by defectors from the former), California is well placed to understand the risks…<\/p>\n

 <\/p>\n

On Monday, the Democratic state passed a law, known as \u201cSB 53,\u201d aimed at regulating the use of artificial intelligence<\/b>.<\/p>\n

 <\/p>\n

Preceded by the European AI Act<\/h3>\n

 <\/p>\n

Presented as unprecedented on a global scale<\/b>, it was nevertheless preceded by similar European legislation, the \u201cAI Act.\u201d<\/b> Already described at the time as pioneering, the latter came into force on August 1, 2024, with gradual implementation starting on February 2, 2025.<\/p>\n

 <\/p>\n

The new Californian law follows two recent attempts across the Atlantic<\/b>, both of which failed.<\/p>\n

 <\/p>\n

On the one hand, the Republican Trump administration attempted to block any regulation of AI by US states, arguing that such regulation would handicap the United States in the AI race against China, which has no such scruples.<\/p>\n

 <\/p>\n

The risk of stifling innovation is not only a concern for Republicans<\/b>. This argument was also used last year by California’s Democratic governor, Gavin Newsom, to veto a bill at the last minute proposed by local senator Scott Wiener, who is affiliated with the same party as him… The bill was an initial attempt to regulate AI, but was deemed too strict.<\/p>\n

 <\/p>\n

A divisive issue<\/h3>\n

 <\/p>\n

The extremely divisive issue caused a stir in Silicon Valley<\/b>, with some leaders speaking out to denounce the risk of an exodus of innovative minds, even though the AI revolution is still in its infancy…<\/p>\n

 <\/p>\n

But this time was the right one, and by enacting SB 53, Governor Gavin Newsom believes he has struck the right balance, helping \u201cthe State of California encourage innovation while protecting public safety.\u201d<\/p>\n

 <\/p>\n

As a reminder, the European legislation<\/b> governing the EU already mentioned also aimed to provide a balanced framework<\/b> for the development of AI on the old continent, allowing it to flourish while putting safeguards in place against potential excesses.<\/p>\n

 <\/p>\n

With AI attracting massive investment and now becoming part of the daily and professional lives of American citizens, Gavin Newsom could not sit idly by.<\/p>\n

 <\/p>\n

\u201cDisturbing progress\u201d<\/h3>\n

 <\/p>\n

\u201cThe own reports of leading AI companies reveal disturbing progress in all categories of threats<\/b>,\u201d warned the think tank set up by Gavin Newsom to review the initial proposal last June. This was a powerful statement coming from experts and academics such as Fei-Fei Li, a Chinese-born scientist who now heads the Stanford Institute for Human-Centered AI and is known as the \u201cgodmother of AI.\u201d<\/p>\n

 <\/p>\n

In its report, the think tank made a number of recommendations for implementing the necessary regulations in this sector without stifling it.<\/p>\n

 <\/p>\n

SB 53 thus imposes transparency requirements<\/b> on companies that develop the most advanced AI models.<\/p>\n

 <\/p>\n

The idea is to require them to report any dangerous misleading behavior detected in AI during testing<\/b>. They will also have to make their security protocols public, report any serious incidents within 15 days, and put in place mechanisms to protect whistleblowers.<\/p>\n

 <\/p>\n

For example, if a model minimizes the brakes designed to prevent it from contributing to the manufacture of biological or nuclear weapons, developers will be required to disclose the incident.<\/p>\n

 <\/p>\n

In general, the law introduces an obligation to report any deceptive or dangerous behavior detected during testing phases.<\/p>\n

 <\/p>\n

The European text is more restrictive in some respects<\/h3>\n

 <\/p>\n

For Senator Scott Wiener, one of the merits of his law is that it makes security protocols public, whereas the European Union’s reference text, although more restrictive in other respects, reserves their transmission to the authorities alone.<\/p>\n

 <\/p>\n

The vote on European legislation also gave rise to heated debates, with some countries, including France, opposing a framework that was too strict<\/b> and likely to stifle AI and hinder its development.<\/p>\n

 <\/p>\n

A compromise was finally reached, with the chosen text imposing constraints proportionate to the risks that AI systems<\/b> pose to society.<\/p>\n

 <\/p>\n

When the risks are limited<\/b>, their authorization in the EU is subject to light transparency requirements<\/b>. However, these requirements are reinforced for systems deployed in high-risk areas such as education, law enforcement, or human resources<\/b>. These requirements include human oversight of the machine, the implementation of a risk management system, and the establishment of technical documentation.<\/p>\n

 <\/p>\n

Specific rules for generative AI<\/h3>\n

 <\/p>\n

Generative AI such as Open AI’s ChatGPT<\/b> is regulated by specific rules<\/b> aimed at ensuring the quality of the data used in the development of algorithms and compliance with copyright laws.<\/p>\n

 <\/p>\n

Sounds, images, and texts generated artificially via AI<\/b> must be identified as such<\/b> to prevent manipulation of public opinion.<\/p>\n

 <\/p>\n

European legislation will not actually apply in its entirety (with a few exceptions) until August 1, 2026, although some provisions have already become mandatory this year.<\/p>\n

 <\/p>\n

Since February 2,<\/b> a few (rare) prohibitions have been implemented, with potential violations punishable by fines of up to 7% of the offender’s global turnover. They prohibit<\/b>, in particular, applications that are contrary to the values cherished by the old continent, such as citizen rating or mass surveillance systems<\/b> similar to those used in China.<\/p>\n

 <\/p>\n

On social network X, Ursula von der Leyen, President of the European Commission, praised \u201ca pioneering European framework for innovative and safe AI\u201d that \u201cwill promote the development of AI that Europeans can trust\u201d and \u201chelp European SMEs and start-ups bring cutting-edge AI solutions to market.\u201d<\/p>\n

 <\/p>\n

European SMEs versus tech giants?<\/h3>\n

 <\/p>\n

Small, virtuous European SMEs and start-ups versus unscrupulous American tech giants<\/b>?<\/p>\n

 <\/p>\n

The reality is certainly less black and white.<\/p>\n

 <\/p>\n

Even before the Californian law came into force, American AI giants, including OpenAI and Anthropic\u2014sensing the wind of change?\u2014had made voluntary commitments to transparency and robustness in their protocols.<\/p>\n

 <\/p>\n

On both sides of the Atlantic, the aim will be to prevent sorcerer’s apprentices from playing with fire with this new technology that poses such a high risk to humanity: AI. And to remember, once again, Rabelais’ valuable warning: \u201cScience without conscience is but the ruin of the soul…\u201d<\/p>\n

 <\/p>\n

Read also > \u201cStarter pack\u201d controversy: artificial intelligence in art criticized once again<\/a><\/p>\n

 <\/p>\n

Featured photo: \u00a9 Getty Images\/Unsplash+<\/p>\n<\/div>","protected":false},"excerpt":{"rendered":"

More than a year after European legislation regulating artificial intelligence came into force, the state of California, the birthplace of this technology, has also just enacted a similar law… While the Trump administration is hostile to any restrictions in the AI race, the initiators of the so-called \u201cSB 53\u201d law have sought to strike the right balance so as not to stifle innovation but to avoid its excesses.<\/p>\n","protected":false},"author":457,"featured_media":35556,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_seopress_titles_title":"%%post_title%% %%sep%% %%sitetitle%%","_seopress_titles_desc":"The state of California has just passed legislation to regulate artificial intelligence (AI).","_seopress_robots_index":"","_seopress_robots_follow":"","_seopress_robots_imageindex":"","_seopress_robots_snippet":"","_seopress_robots_primary_cat":"none","_seopress_robots_breadcrumbs":"","_seopress_robots_freeze_modified_date":"","_seopress_robots_custom_modified_date":"","_seopress_robots_canonical":"","_seopress_social_fb_title":"","_seopress_social_fb_desc":"","_seopress_social_fb_img":"","_seopress_social_fb_img_attachment_id":0,"_seopress_social_fb_img_width":0,"_seopress_social_fb_img_height":0,"_seopress_social_twitter_title":"","_seopress_social_twitter_desc":"","_seopress_social_twitter_img":"","_seopress_social_twitter_img_attachment_id":0,"_seopress_social_twitter_img_width":0,"_seopress_social_twitter_img_height":0,"_seopress_redirections_value":"","_seopress_redirections_enabled":"","_seopress_redirections_enabled_regex":"","_seopress_redirections_logged_status":"both","_seopress_redirections_param":"","_seopress_redirections_type":301,"_seopress_analysis_target_kw":"","_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":""},"categories":[21],"tags":[432],"class_list":["post-35559","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-business-en","tag-artificial-intelligence"],"_links":{"self":[{"href":"https:\/\/magazine.luxus-plus.com\/en\/wp-json\/wp\/v2\/posts\/35559","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/magazine.luxus-plus.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/magazine.luxus-plus.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/magazine.luxus-plus.com\/en\/wp-json\/wp\/v2\/users\/457"}],"replies":[{"embeddable":true,"href":"https:\/\/magazine.luxus-plus.com\/en\/wp-json\/wp\/v2\/comments?post=35559"}],"version-history":[{"count":0,"href":"https:\/\/magazine.luxus-plus.com\/en\/wp-json\/wp\/v2\/posts\/35559\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/magazine.luxus-plus.com\/en\/wp-json\/wp\/v2\/media\/35556"}],"wp:attachment":[{"href":"https:\/\/magazine.luxus-plus.com\/en\/wp-json\/wp\/v2\/media?parent=35559"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/magazine.luxus-plus.com\/en\/wp-json\/wp\/v2\/categories?post=35559"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/magazine.luxus-plus.com\/en\/wp-json\/wp\/v2\/tags?post=35559"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}