Cybersecurity Snapshot: Cyber Pros Emerge as Bold AI Adopters, While AI Changes Data Security Game, CSA Reports Say

A weekly roundup highlights that cybersecurity teams are rapidly adopting AI tools (per a CSA–Google survey), CSA proposes new AI data security controls to address prompt injection, model inversion, and multimodal leakage, and PwC expects responsible AI practices to operationalize by 2026; FinCEN reports nearly 4,200 ransomware incidents and $2.1B in payments from 2022–2024 with Akira, ALPHV/BlackCat, LockBit, Phobos, and Black Basta prevalent; and the UK NCSC stresses that prompt injection is distinct from SQL injection and calls for mitigations such as least‑privilege design, developer training, and monitoring.