Podcast Examines Hacked Supermarket Refrigeration, AI-Assisted Ransomware and Mythos
· 29:28 · Watch on YouTube ↗
Summary
IBM’s Security Intelligence panel discusses a French supermarket refrigeration-system attack linked to outdated firmware, AI-related ransomware findings from a Semperis survey, and Bank of America’s use of Anthropic Mythos to find vulnerabilities.
Key points
- Hackers took control of refrigeration management systems from Danfoss, Copeland and Dixell at French supermarket chains; some fridges reportedly overheated, smoked or caught fire.
- Danfoss issued a firmware update after Team82 discovered vulnerabilities, but many refrigeration units were still running outdated firmware.
- The panel discusses weighing planned downtime for OT patching against the risk of prolonged outages, product loss and physical safety impacts.
- The 2026 Semperis Ransomware Risk Report, based on a survey of 1,200 IT professionals across nine countries and eight industries, reports that 84% of ransomware attacks involved AI assistance and 44% involved exploitation of organizations’ AI agents or systems.
- Panelists recommend continuous governance, monitoring and safeguards for AI-agent identities and third-party agent access, alongside automated security controls and resilience planning.
- The podcast says Bank of America hired 1,000 contractors to help patch hundreds of vulnerabilities found with Anthropic Mythos through the Project Glass Wing initiative; panelists stress ongoing patching and controlled contractor access.
Article Details
- Topics
- French supermarket refrigeration-system cyberattack
- Outdated firmware and OT patching
- AI-assisted ransomware and survey findings
- AI-agent identity risks and governance
- Mythos-assisted vulnerability discovery
- Contractor access management and remediation
People
Claire NunezIBM's X-Force Cyber Range Creative Director and podcast panelist; discussed OT patching, defensive automation and safe AI-agent use.Courtney GussSemperis Associate Vice President of Crisis Management, quoted on the risk that AI agents could compromise company identity systems.M.J. VaidyaGlobal Senior Partner, Cyber Strategy and Risk, and podcast panelist; discussed product security, AI governance, resilience and access controls.Mr. MoynihanReferenced as discussing patching and remediation in connection with Bank of America's Mythos-related vulnerability findings; the transcript does not provide his full name or title.