AI accountability is moving from theory to execution, as organizations increasingly recognise the need for robust governance structures and chains of ownership, according to new research from BSI.
North America
The latest red team research from Barracuda shows how, in a controlled simulation, researchers took over a Copilot-enabled account and used the assistant against its own organization.
Detection speed has often been used as the ultimate metric of cyber preparedness. However, new research from ManageEngine reveals an uncomfortable truth: catching an attacker quickly does not guarantee a swift recovery. VimalRaj explores why the gap between detection and recovery is widening.
Every organization tells stories about itself. Those narratives can create genuine resilience capabilities – but they can also conceal blind spots, reinforce fragility, and restrict adaptation.
Travel safety apps cannot protect employees they cannot locate. Matt Atkinson examines the hidden connectivity dependency in travel risk management and the potential for mobile-network data to provide a resilient fallback.
Zero Networks has announced Least Agency Enforcement, a new capability that applies the OWASP principle of ‘least agency’ to help organizations safely deploy AI agents.
According to IBM’s 2026 Cost of a Data Breach Report, one in four malicious breaches was AI-enabled during the 12-month period covered by the report.
A new report from Ivalua has found that, in the past 12 months, 70% of businesses have had between one and five critical suppliers fail, while 41% say they’re one supplier failure away from a supply chain crisis.
Security teams remain overly committed to keeping humans in the loop and it isn’t a sustainable approach says Ashley Leonard.
Cyber security agencies from Australia, the US, the United Kingdom, and Canada have published new guidance to help critical infrastructure operators isolate vital operational technology (OT) and enabling systems during a major cyber incident or geopolitical crisis.