LLaMA 4 leak: What Meta didn't prepare for
Elena Vance
23 Apr 2026·13m·💬 0
30 articles
AIThe 2025 LLaMA 4 leak exposes critical security vulnerabilities in Meta's open-source AI, raising questions about model safety.
AIDeepSeek R2 emerges with near-parity to GPT-4o, raising US export control debates. The model's rapid release shocks experts.
AIAnthropic's new policy of reporting AI vulnerabilities directly to the US government creates a legal and ethical chasm, raising critical questions about AI ethics in 2025.
AIThe UK competition regulator just launched a formal probe into the AI partnerships of Microsoft and Amazon, a move with global antitrust implications in 2024.
AIPerplexity's new Pro Search agent directly challenges Google's core business by rewriting the economics and architecture of web discovery.
AIMicrosoft's AI-powered Recall feature is a bonanza for data regulators, creating an unprecedented legal liability.