Real-time visibility, safety, and security for your GenAI-powered agents and applications
Proactively test GenAI models, agents, and applications before attackers or users do
Deploy generative AI applications and agents in a safe, secure, and scalable way with guardrails.
Proactively identify vulnerabilities through red teaming to produce safe, secure, and reliable models.
Signed just this month, the California Age-Appropriate Design Code Act (“ADCA”) places responsibilities on businesses that provide a product or service likely to be accessed by children. The law will be adopted on July 1, 2024, placing new obligations that Trust & Safety teams should begin preparing for now.
ActiveFence spoke with Michal Brand – Gold, our General Counsel VP, to understand who the act will apply to, what the obligations are, and how companies can prepare.
The ADCA seeks to protect children’s data. Specifically, the act aims to mitigate the following risks:
The ADCA will apply to businesses that provide an online service, product, or feature likely to be accessed by children in California. These companies:
Platforms that the ACDA applies to must comply with the following requirements:
The CA Age-Appropriate Design Code can hold violators liable for a fine of up to $2,500 per affected child for each negligent violation and up to $7,500 per affected child for each intentional violation.
However, businesses that comply with the law will be given notice before any actions are initiated. The business will have 90 days to rectify any violations before penalties are given.
An essential piece of the ADCA is creating a working group whose purpose is to develop best practices for implementing the law, as well as identify services that are required to comply with it.
Additionally, the working group has been given access to leverage California’s Privacy Protection Agency (CPPA) which has years of experience developing data privacy policies.
Given that the law establishes a Working Group and previously established the CPPA, it seems likely that the ADCA will be enforced.
The California act is modeled after the UK’s Age Appropriate Design Code which took effect in September 2021. Given their similar requirements, many major tech companies have already implemented measures to meet the UK’s code and, therefore, California’s requirements.
Major tech companies redesigned online platforms to comply with the Children’s Code. However, we haven’t yet seen enforcement or fines from the code.
Instead, the focus of the UK’s code authority, the ICO, has been to help companies find solutions. As such, the ICO issues design guidance, a self-assessment risk tool, and transparency best practices.
In the evolving legal landscape of online liability, preparedness is key for online platforms to stay compliant. To do so, teams must remain up to date on legislation worldwide that affects the internet. Check out our Trust & Safety Compliance Center and help ensure your platform is compliant.
Prompt injection, memory attacks, and encoded exploits are just the start. Discover the most common GenAI attack vectors and how red teaming helps stop them.
The EU has formally sanctioned key players behind Russia’s coordinated disinformation ecosystem. These campaigns, long monitored by ActiveFence, reveal a complex strategy built on narrative laundering, infrastructure resilience, and long-term influence.
AI misuse isn’t hypothetical - it’s happening now. This blog introduces ActiveFence’s latest guide for operationalizing AI safety and security with six real-world strategies to move from principle to protection.