Cartwright Act: violations.

Bill was formally enacted as Chapter 338 with the Governor's approval date of October 6, 2025 added to the header.

Document formatting and line breaks were adjusted throughout to match the chaptered statutory publication layout; no statutory language was altered.

The document changed from an official chaptered legislative format (with authentication headers, page numbers, and formal layout) to a web page scrape that includes site navigation elements (skip to content, home, accessibility, FAQ, feedback, sitemap, login, search fields, bill version history links, etc.) surrounding the identical bill text.

Header changed from amendment history listing to enrolled bill format with passage dates and signature blocks for both chambers and the Governor.

Line breaks and page numbering were reformatted throughout the document (page 92 became page 91) but no words in the statutory text were added, removed, or altered.

The definition of 'common pricing algorithm' was narrowed: removed 'process or rule,' removed the reference to 'machine learning or other artificial intelligence techniques,' simplified 'processes the same or substantially similar uses competitor data' to 'uses competitor data,' and removed the trailing phrase 'using the same or performing a substantially similar function.' The old version also contained apparent drafting errors with repetitive language.

The coercion provision in Section 16729(b) now requires that the price or commercial term be 'recommended by the common pricing algorithm' before it constitutes a violation. Previously, coercing someone to set or adopt any recommended price or commercial term was sufficient.

Section 16729(b) was simplified from a complex two-prong test (distributing to two or more persons with intent plus coercion, OR using the algorithm while knowing/should know about a price-fixing scheme or coercing) down to a single standard: it is unlawful if the person coerces another person to set or adopt a recommended price or commercial term. This removes the 'knows or should know' mens rea requirement and the multi-person distribution element from the standalone prohibition.

The definition of 'common pricing algorithm' in Section 16729(d)(3) was expanded to include 'methodology,' 'computer, software, or other technology,' added a requirement that it be 'used by two or more persons,' changed the data reference to 'competitor data,' and broadened the covered actions from 'recommend or set' to 'recommend, align, stabilize, set, or otherwise influence' a price or commercial term.

The core legal mechanism changed from establishing presumptions of antitrust violations (contracts in restraint of trade under Chapter 1 or Chapter 4) to a direct prohibition making it 'unlawful' to use or distribute qualifying pricing algorithms. This removes the need to prove a separate antitrust violation and creates a standalone offense.

The safe harbor defense in subdivision (c) was narrowed: a defendant no longer needs to show they lacked knowledge that the algorithm 'used nonpublic competitor data' - they only need to show they did not know the algorithm 'was used by two or more persons to set or recommend a price or commercial term.' This makes the defense easier to satisfy.

The definition of 'nonpublic competitor data' removed an exclusion that had carved out 'information distributed, reported, or otherwise communicated in a way that does not reveal any underlying data from a competitor, such as narrative industry reports, news reports, business commentaries, or generalized industry survey results.' This broadens what counts as nonpublic competitor data.

Joint and several liability for distributors was narrowed from covering 'any violation of this chapter or Chapter 4 (commencing with 17000)' to only 'any violation of this section,' limiting the scope of potential liability exposure.

New prohibition making it unlawful to use or distribute a pricing algorithm that uses data other than nonpublic competitor data (i.e., public data, one's own data) if the person distributes the algorithm to two or more persons in the same market or uses it to set prices and another person also uses it in the same or related market. This expands coverage well beyond the original bill which only targeted algorithms using nonpublic competitor data.

The conditions for presumptions of antitrust violation now include a pricing algorithm trained on substantially similar data to the one at issue, not just the exact same algorithm. This makes it easier to establish liability when similar algorithms are in play across multiple market participants.

The definition of 'nonpublic competitor data' was clarified to explicitly state it includes aggregated data (adding 'including, but not limited to, aggregated data'), removing any ambiguity about whether aggregated competitor information falls within the prohibition.

Several subsections were renumbered and verb tenses were updated from past to present (e.g., 'distributed' to 'distributes,' 'knew' to 'knows'), reflecting drafting refinements rather than substantive policy changes.

The standalone prohibition on using or distributing any pricing algorithm that uses, incorporates, or was trained with nonpublic competitor data (old Section 16729(a)(1)) is entirely removed, along with the definitions of 'nonpublic competitor data' and 'nonpublic data.'

A new prohibition is added making it unlawful to use or distribute a common pricing algorithm 'as part of a contract, combination in the form of a trust, or conspiracy to restrain trade or commerce,' directly linking algorithm use to existing antitrust violations.

The conditions for liability when distributing a common pricing algorithm now require both distribution to two or more persons AND coercion of any person to set or adopt a recommended price, rather than merely intent that the algorithm be used in the same market.

The conditions for liability when using a common pricing algorithm now require either knowledge that another person is adhering to or participating in a scheme to fix prices, or coercion of any person to set or adopt a recommended price, rather than merely knowing the algorithm was used by another person in the same market.

The safe harbor defense (old subdivision (c)) that allowed a person who did not develop or distribute the algorithm to avoid liability by proving by clear and convincing evidence they lacked actual knowledge is removed entirely.

Joint and several liability is simplified: the old version required proof that the distributor knew or could have reasonably known about nonpublic competitor data or multi-party use; the new version simply states any person who violates the section is jointly and severally liable.

The definition of 'commercial term' no longer includes 'rebates or discounts made available,' narrowing what qualifies as a commercial term under the bill.

The separate definition of 'pricing algorithm' (any computational process that processes data to recommend or set a price) is removed, leaving only the definition of 'common pricing algorithm.'

A new definition of 'common pricing algorithm' is added: any process or rule (including machine learning or AI) that processes the same or substantially similar data to recommend or set a price or commercial term using the same or performing a substantially similar function. The prohibition in subdivision (b) now applies only to 'common pricing algorithms' rather than all pricing algorithms.

The definition of 'antitrust laws' is narrowed from referencing the Clayton Act, Section 45 of Title 15 USC, and the Cartwright Act to simply 'the provisions of this part' (California's Business and Professions Code antitrust chapter).

The definition of 'commercial term' changes from an exhaustive list ('means any of the following') to a non-exhaustive one ('means includes, without limitation, any of the following').

The geographic scope is narrowed from 'the same market or a related market' (no geographic limit) to use 'in the jurisdiction of this state.' Both prongs of the prohibition in subdivision (b) now require the algorithm to be used for products or services in California.

Assembly Member Ward is added as a coauthor of the bill.

The definition of 'person' is expanded to explicitly exclude 'the end consumer of a product or service,' meaning individual consumers who use pricing algorithms cannot be held liable under this section.

Subdivision (b)(2) now requires that the person 'knows or should know' the common pricing algorithm is, was, or will be used by another person to set prices in California. The prior version only required that the algorithm actually was used by another person, without a knowledge element for the user.

The joint and several liability provision (subdivision d) now requires that the distributing person 'who violates this section' is jointly and severally liable, rather than simply being liable for any violation. This adds an additional requirement that the distributor must themselves be in violation.