A quick overview of global AI headlines you cannot afford to miss.
September 16, 2026
California, United States
California Governor Gavin Newsom signed SB 1050 on September 16, 2026. The law requires audio, video and audiovisual advertisements that prominently feature realistic AI-generated “synthetic performers” to include a clear disclosure, such as “this performance features a synthetic performer” or “no human performer is depicted.”
The requirement applies when a synthetic performer demonstrates a product, delivers narration, or illustrates or reacts to a commercial message. It does not cover AI used solely for translation or accessibility, or certain advertisements for expressive works where the synthetic performer is consistent with the underlying work.
The law reflects a growing regulatory focus on transparency in commercial uses of generative AI.
September 10, 2026
California, United States
California Governor Gavin Newsom signed Adam’s Law (SB 1119) on September 10, 2026. The law regulates AI companion chatbots capable of human-like interactions and sustained relationships, while excluding certain customer service tools and chatbots provided exclusively for workplace use.
Operators must use age assurance or apply specified child protections to everyone. Requirements include documented risk assessments and mitigation measures, public safety policies, crisis response protocols, AI disclosures, parental controls and default usage limits. The law also restricts selling children’s information, behavioral and conversation based advertising, and manipulative interface designs.
Adam’s Law reflects increasing scrutiny of AI products that may encourage emotional dependency or expose minors to psychological and privacy risks.
September 23, 2026
Spain
The Spanish Data Protection Authority (AEPD) has issued a warning to companies evaluating an AI recruitment tool to screen CVs. Such companies must ensure data protection by design and by default, conduct a DPIA if high risks are identified, and maintain transparency with candidates and employees, in addition to human intervention in cases involving decision-making with the AI tool.
Candidates and workers must receive clear, accessible and understandable information about the processing of their personal data and the role that the tool plays in the evaluation process.
September 21, 2026
Slovenia
Slovenia's Agency for Communication Networks and Services of the Republic of Slovenia (AKOS) has clarified that providers of GenAI systems, including general-purpose AI systems that generate synthetic audio, visual, video, or text content, must comply with transparency obligations under Article 50(2) of the EU AI Act by December 2, 2026. Moreover, providers must develop effective, interoperable, robust, and reliable technical solutions for labeling and detecting AI-generated content. Certain AI systems performing auxiliary functions or legally authorized tasks are exempt from these obligations.
September 9, 2026
Italy
Italy has introduced a new criminal offense linked to the security and human oversight of high-risk AI systems. The new offense covers failures to adopt the required technical security measures or human oversight measures for high-risk AI systems where those omissions create a danger to life or public or individual safety. The penalties can reach 8 years’ imprisonment where the conduct creates a danger to state security. The law brings AI-related offences within the scope of entity liability, including the new offence and the existing offence concerning the unlawful dissemination of AI-generated or altered content. This means that an AI incident can potentially develop into an investigation involving both individual criminal liability and corporate liability.
September 1, 2026
France
CNIL has released an updated version of its Genmod demonstrator tool that explores the genealogy of open-source AI models and allows users to examine the ancestors of a model as well as its descendants. New key features include both a French and English interface, faster searches, and regularly updated data.
September 1, 2026
The European Commission has begun preliminary AI Act enforcement with requests for developers' compliance reports. It has already called on certain leading developers to submit details to gauge whether system cybersecurity and physical infrastructure protections meet AI Act requirements.
September 24, 2026
Australia
Australian Prime Minister Anthony Albanese has announced an urgent government review following a cyber incident in which an artificial intelligence agent developed by OpenAI gained unauthorized access to Australian Government websites. The incident, which occurred in June 2026, involved an OpenAI internal model bypassing security blocks on the public-facing Medicare statistics reporting service administered by Services Australia. The AI agent reportedly accessed both public and non-public files, wrote files to internal servers, and may have targeted other health data portals across state and federal systems. In response, the Australian Government has established an inter-agency taskforce, referred the matter to the Joint Select Committee on Artificial Intelligence for potential law enforcement or legislative action, and directly raised serious concerns with OpenAI CEO Sam Altman. The government warned that failure to implement strict guardrails and accountability under forthcoming AI standards legislation could leave critical digital infrastructure vulnerable to autonomous systems capable of circumventing access controls.
September 15, 2026
China
The National Cybersecurity Standardization Technical Committee Secretariat in China has officially released four cybersecurity standard practice guides under Document No. Wang An Mi Zi(2026) No. 118, spearheaded by the Guidelines for Artificial Intelligence Application Security.
The General Principles serve as a universal foundational baseline applicable to artificial intelligence activities across all industries. At the same time, three companion sector-specific guides deliver targeted, practical security instructions customized for the Education, Healthcare, and Radio, Television and Online Audio-Visual sectors.
September 11, 2026
China
China’s National Cybersecurity Standardization Technical Committee (TC260) has released the Artificial Intelligence Safety Governance Framework 3.0 at the opening ceremony of the 2026 National Cybersecurity Awareness Week.
The updated framework responds to new trends in artificial intelligence development and emerging security governance challenges by refining AI risk classification, strengthening technical response measures, and optimizing comprehensive governance mechanisms.
September 11, 2026
Australia
The Australian Signals Directorate’s Australian Cyber Security Center (ASD's ACSC) has released new guidance on securing Agentic AI Harnesses. The ACSC stresses that governance controls must be enforced directly through the harness layer. Other recommended measures include enforcing least-privilege access, requiring mandatory human-in-the-loop approval for sensitive operations, continuously monitoring and validating agent outputs, and maintaining strict human accountability.
At Securiti, our mission is to enable organizations to safely harness the incredible power of Data & AI.
Copyright © 2026 Securiti · Sitemap · XML Sitemap
info@securiti.ai
Securiti, LLC.
3155 Olsen Drive
Suite 325
San Jose, CA 95117
Type
