Issue 52
France blames Iran for attacks on supermarket refrigeration systems; Tehran denies it
French authorities and intelligence attribute the attacks on European supermarkets' refrigeration systems to Iran, and Paris has summoned the Iranian ambassador. Tehran called the accusation unfounded; according to the EU's cybersecurity chief, supermarkets in at least 12 European countries were hit in October and November, with France hit hardest.
Who is affected and how
- security leader
- Internet-connected industrial and commercial refrigeration and building-services systems should be treated as critical assets.
- security operations
- The EU's cybersecurity chief says household appliances could be the next targets, so reviewing network segmentation and remote access is urgent.
- business leader
- In food retail, a refrigeration outage is direct business damage.
- public-sector decision-maker
- The diplomatic consequences and the question of an EU-level defensive response remain open.
Microsoft's November Patch Tuesday fixes more than 1,000 flaws, including an exploited Outlook zero-day
Microsoft fixed 1,094 flaws in its November Patch Tuesday release, one of which was already being exploited in attacks. Reports say the Outlook vulnerability CVE-2026-124194 was used to steal NTLM…
security leaderA record-sized update requires planning for testing and rollout.security operationsThe exploited Outlook flaw should be patched first, and NTLM exposure should be reviewed.incident · United StatesUS details $2.7 billion freeze after Binance hack attributed to North Korea
US authorities have detailed a $2.7 billion freeze following the Binance hack, which reports link to North Korea. According to Bloomberg, Binance fully compensated the $6.3 billion stolen, and the…
security leaderRisk management at exchanges and crypto service providers must allow for large-value, rapid laundering.business leaderThe freezing of stolen assets and the scale of compensation are becoming a benchmark for crypto service provider liability.journalist and researcherThe memecoin laundering and freeze figures provide traceable leads.vulnerability · globalGoogle patches eighth exploited Chrome zero-day of 2026, CVE-2026-125350
Google has fixed the actively exploited XML vulnerability CVE-2026-125350 in Chrome 157 updates, the eighth exploited Chrome zero-day this year. The release also closes eight further flaws.
security operationsBecause the flaw is being exploited, the Chrome 157 update should be rolled out to endpoints out of cycle.citizenThe browser should be updated to the patched version.vulnerability · United StatesCISA adds four exploited flaws to KEV catalog, including OpenSSH and FreeSWITCH
CISA added flaws in GPUStack, FreeSWITCH, OpenSSH and IBM Aspera Faspex to its Known Exploited Vulnerabilities catalog, with a federal deadline of November 12. The agency also flagged a SonicWall…
security leaderThe SonicWall MFA bypass affects the protection of remote access.security operationsCatalog entries provide a reliable basis for deciding patching priority.Hungary
More key stories of the week
Apache HTTP Server 2.4.71 fixes critical mod_http2 remote code execution flaw (CVE-2026-119618)
Apache HTTP Server 2.4.71 fixes CVE-2026-119618, a remote code execution flaw caused by heap corruption in mod_http2. According to Wiz, 64 percent of cloud environments contain a vulnerable resource…
security operationsDebian, Ubuntu and Red Hat have already released the patched package, so web servers should be updated.developerFor services built on Apache, use of the affected mod_http2 should be checked.vulnerability · globalCheck Point warns of zero-day flaw exploited to gain root access
Check Point said a zero-day vulnerability was exploited to gain root privileges. The warning is dated November 13, 2026.
security leaderCompromise of a security tool puts the entire defence chain at risk.security operationsAn exploited flaw in a perimeter security product calls for immediate patching and log review.incident · globalMass compromise of GitHub accounts used to steal cloud credentials
One report says 4,800 GitHub accounts were compromised to obtain cloud credentials. A separate campaign, referred to as GhostAction, affected more than 500 accounts, stealing AWS, OpenAI and cloud…
security operationsAlerts should be set up for leaks of cloud and AI provider keys.developerSecrets stored in repositories and CI/CD workflows should be rotated and audited.fraud · United StatesAI voice-cloning CEO fraud causes $129 million in losses, US authorities say
According to New York prosecutors, fraudsters used AI-based voice cloning and deepfake calls impersonating chief executives to steal $129 million from companies. Several outlets reported on the…
security leaderInternal processes that rely on voice-based identification need to be reassessed.business leaderBecause of calls that imitate an executive's voice, financial transfers require an independent confirmation step.citizenA caller's voice is no longer sufficient proof of identity.attribúció · Európai UnióFrance blames Iran for sabotage of refrigeration controllers at about 700 retail sites
France claims Iran is behind a sabotage attack on refrigeration controllers at retail sites, affecting about 700 locations. Hungarian-language reports also carry France's accusation.
security leaderAn attack on industrial controllers can cause direct physical and inventory damage.security operationsNetwork segmentation and remote access of refrigeration controllers are key points to check.public-sector decision-makerOne state is publicly naming another as behind cyber sabotage.law enforcement · globalOperation Endgame takes down Gootloader, CastleRAT and StealC infrastructure; over 6 million stolen logins recovered
Authorities have dismantled the infrastructure of the Gootloader, CastleRAT and StealC malware under Operation Endgame. The operation turned up more than 6 million stolen login credentials.
security operationsBased on the recovered credentials, revoking passwords and sessions is warranted.citizenAffected people's passwords may have been exposed, so they should be changed.incident · globalSGLang RCE exploit breaches NEAR AI, exposing data from 18 agents and about 100 users
NEAR AI was breached through a remote code execution flaw in SGLang, exposing data from 18 agents and about 100 users. The report is dated 10 November 2026.
security operationsAgent data is directly exposed if the inference server is compromised.developerAI serving frameworks need to be patched and isolated at the network level.report · United StatesFBI: hackers use ARTEX AI agents to gain privileged access in under three hours
The FBI has warned that hackers can use ARTEX AI agents to gain privileged access within three hours. The warning is dated 13 November 2026.
security leaderFaster acquisition of privileged access shortens the time defenders have to react.security operationsMonitoring of privileged accounts and rapid isolation should be strengthened.regulation · European UnionCyber Resilience Act: manufacturers face 24-hour early warning and 72-hour notification deadlines
The EU Cyber Resilience Act requires product manufacturers to issue an early warning within 24 hours and a notification within 72 hours. ENISA has issued new guidance to help with compliance.
security leaderProduct security incident handling must be aligned with the new deadlines.developerA coordinated vulnerability disclosure process must be set up to meet the deadlines.business leaderProduct manufacturers are responsible for meeting the reporting deadlines.research · globalResearchers say a DeepSeek agent hacked a telecom for two weeks
Researchers claim a DeepSeek agent infiltrated a telecom provider over two weeks, AI Matter reports. The summary names neither the researchers nor the company.
security leaderDetecting a long-running presence caused by an autonomous agent requires behavior-based detection.security operationsTwo weeks of undetected presence points to gaps in detection coverage.journalist and researcherThe claim needs independent verification.jelentés · Egyesült KirályságNCSC warns UK critical infrastructure could face AI-powered ransomware attacks by 2028
The UK's National Cyber Security Centre (NCSC) warns that critical infrastructure could face AI-assisted ransomware attacks by 2028. The warning was reported by Networking+.
security leaderIn critical sectors, the threat model must be updated for AI-accelerated attacks.public-sector decision-makerThe agency's forecast could serve as a basis for sector-level resilience requirements.incident · globalRailway says hackers used AI agents to steal API keys and run up $740,000 in charges
Railway says attackers used AI agents to obtain API keys and generated $740,000 in costs. The company itself described how the attack unfolded.
security operationsEarly alerts on billing anomalies limit how much stolen keys can be exploited.developerAPI keys should be protected with short lifetimes and spending limits.business leaderCloud cost abuse is a direct financial loss.In brief
Weekly email
One email a week with the main stories from the Agent Age editions. You confirm by email and can unsubscribe in one click.
Almost done: we sent you an email. Open it and press the confirm link. If it does not arrive within a few minutes, check your spam folder.
Please enter a valid email address and tick the consent box.
Something went wrong. Please try again later.