AI & software security
-
Databricks deal pushes security lakehouse model
Databricks has agreed to acquire Panther, extending its push into security operations as AI, data scale, and SIEM costs reshape enterprise detection and response.
-
Copilot flaw exposes AI data risk
Varonis says Microsoft patched a critical Copilot vulnerability chain that could have allowed sensitive Microsoft 365 data to be extracted through a crafted link.
-
ENISA meeting puts AI access in focus
ENISA’s planned meeting with Anthropic comes as European cyber agencies confront access, assurance, and dependency issues around powerful AI models used in security work.
-
Databricks moves further into security data
Databricks’ planned acquisition of Panther reflects enterprise pressure to consolidate security telemetry, automate investigation, and rethink legacy SIEM economics.
-
Chip security meets post-quantum design
Agile Analog and Xiphera are combining anti-tamper semiconductor IP with post-quantum cryptographic IP for long-life secure chip designs.
-
Malicious plugins put AI keys at risk
A JetBrains Marketplace campaign shows how developer plugins, AI assistants, and API keys can become part of the software supply chain risk surface.
-
FCA brings frontier AI into resilience planning
The FCA has directed firms towards CMORG guidance linking frontier AI capability with cyber security, vulnerability handling, and operational resilience in financial services.
-
Europe confronts frontier AI dependency
Anthropic’s model access restrictions have placed frontier AI inside Europe’s debate over cyber capability, export controls, resilience, and digital sovereignty.
-
Anthropic suspension exposes AI dependency
A US directive forced Anthropic to suspend access to Fable 5 and Mythos 5 for foreign nationals, exposing dependency risk around frontier AI security tools.
-
UK tests AI in cyber defence
A UK government pilot found 407 security findings across nine organisations using frontier AI models, showing defensive potential and governance pressure.