Leading AI developers call for slowing the pace of development due to loss-of-control risks.
Anthropic CEO Dario Amodei called for slowing down development across the sector and involving independent auditors, citing the safety risks of advancing AI models. The proposal, backed by executives from OpenAI, Google DeepMind, and xAI, received mixed reactions from politicians due to global tech competition between China and the US. While lawmakers and the White House in the US remain divided on balancing innovation and safety, efforts to establish global safety standards led by the UK continue.
Click on the link for the full story.
Lack of multilingual safety and evaluation in AI systems brought to the global agenda following New Delhi Summit.
At the New Delhi AI Impact Summit, 13 leading AI companies committed to strengthening multilingual evaluations, highlighting a critical gap in global governance. Current safety testing, content moderation, and risk assessments are conducted almost entirely in English, significantly increasing security vulnerabilities and breach risks in other languages. Experts recommend that organizations immediately implement compliance processes, localized transparency principles, and multilingual auditing mechanisms during this period of global AI regulation.
Click on the link for the full story.
The rise of autonomous AI agents renders traditional data protection laws and consent mechanisms inadequate.
Autonomous AI agents capable of setting their own goals and making dynamic decisions challenge fundamental principles underlying traditional data protection regulations, such as the EU General Data Protection Regulation ("GDPR"). The unpredictable operation of these systems and their continuous reprocessing of data for different purposes make explicit consent prior to data collection, purpose limitation, and data segregation practically impossible. Experts suggest developing dynamic governance models and local data processing architectures instead of static rules.
Click on the link for the full story.
European Commission President presents new digital vision for frontier AI oversight and limits on children's social media access.
In her State of the Union address, European Union President Ursula von der Leyen outlined plans to slow uncontrolled technological advancement and curb cyber risks by collaborating with frontier AI developers. Prioritizing the integration of AI into five strategic sectors—including healthcare and defense—the Commission announced the draft "EU KIDS Act," which includes a social media ban for children under 13 for online safety. Additionally, member states were called upon to avoid imposing additional administrative burdens beyond EU legislation.
Click on the link for the full story.
Privacy-Enhancing Technologies create new value for business and research while protecting sensitive data.
At a web conference organized by the IAPP, it was emphasized that Privacy-Enhancing Technologies (PETs)—such as differential privacy, fully homomorphic encryption, and federated learning—are critical tools not only for legal compliance but also for unlocking data value. Sharing examples from organizations such as Google, Mastercard, and the US Census Bureau, speakers noted that data is securely shared in financial fraud prevention, traffic analysis, and pediatric cancer research, underscoring the need to select the right technology in a balanced manner based on specific needs.
Click on the link for the full story.
European Commission announces draft KIDS Act restricting children's social media access and imposing strict duties on platforms.
The European Commission officially announced the draft KIDS Act (EU KIDS Act), which prohibits children under 13 from creating social media accounts and provides limited access under parental supervision for the 13–14 age group. While requiring large platforms to undergo independent audits, remove addictive designs, and integrate age verification tools, the law foresees fines of up to 6% of global annual turnover in cases of non-compliance. Meanwhile, experts and industry representatives warn that age verification processes could pose privacy risks.
Click on the link for the full story.
UK High Court delivers critical decision regarding the scope and application of data subject access requests.
Deciding the Ashley v. HMRC case, the UK High Court confirmed that the right of access to data under Article 15 of the UK GDPR is not a right to receive entire documents, but rather a right to receive personal data. The ruling emphasized that data controllers cannot narrow the search scope by citing internal organizational structures, and that where raw data presentation renders the text incomprehensible, controllers must also provide necessary contextual information or document excerpts to ensure the intelligibility of the data.
Click on the link for the full story.
Chinese and Hong Kong authorities step up oversight by issuing new guidelines on AI and data security.
While the Supreme People's Court of China issued its first guiding principle regarding AI lawsuits, National Technical Committee 260 published Version 3 of the Security Governance Framework covering AI agents. The Hong Kong Privacy Commissioner for Personal Data announced privacy recommendations tailored to autonomous AI systems, as regulatory bodies across both regions tightened sanctions against data breaches, unauthorized cross-border transfers, and biometric data collection violations, raising compliance expectations for businesses.
Click on the link for the full story.
Public concerns regarding AI safety and governance move to the top of the political agenda in the US.
The fact that a majority of US voters view the existential risks of AI as a serious threat is forcing political parties to abandon policies of inaction. While legislative proposals introducing pre-deployment auditing and duty-of-care obligations for frontier models gain traction among both Democrats and Republicans in Congress, the White House and certain circles argue that existing product liability laws are sufficient so as not to stifle innovation. These differences of opinion signal that AI will be one of the most critical debate topics in the upcoming election cycle.
Click on the link for the full story.
Italian Data Protection Authority sanctions media company for publishing unauthorized AI-generated deepfake video.
The Italian Data Protection Authority ruled that media company RTI violated the GDPR by broadcasting an unauthorized AI-generated deepfake video of a famous journalist on a television show. The authority held that inadequate disclaimers, a lack of transparency across distribution channels, and the non-obvious nature of the satirical intent caused the footage to be perceived as authentic, thereby damaging the individual's reputation. Due to violations of transparency and data protection by design principles, the media company was issued a warning, and all further processing activities for the videos in question were banned.
Click on the link for the full story.