Collection of cybersecurity reports, surveys, research papers, market studies, and industry intelligence.
725 articles
Apiiro researchers found 2,539 secrets across 1,994 public repositories containing Claude Code configuration files. They did not test credentials; one organization confirmed an exposed key was live.
The ITRC’s 2026 report finds unauthorized device access rose to 27.2% of reported identity compromises, while 25.6% of individuals contacting the center faced multiple incidents. Its findings describe patterns in identity misuse, recovery and prevention.
Ping Identity and Deloitte outline six 2026 IAM trends, including AI-agent identity, helpdesk and hiring fraud, decentralized identity, quantum-resistant cryptography and continuous verification, alongside global trust survey findings.
Google Cloud’s H1 2026 Threat Horizons report says identity issues featured in 83% of H2 2025 cloud/SaaS incidents and data was targeted in 73%, while software exploitation overtook weak credentials; it details insider risks, intrusion cases and response guidance.
HID’s 2026 survey of more than 1,500 end users and industry partners finds identity management and physical-digital integration are priorities, while fragmented systems, integration costs, and biometric privacy concerns remain major obstacles.
A survey of 254 U.S. organizations found a gap between self-assessed file security maturity and evidence from respondents’ answers. The report also finds rising concern about AI-driven file threats and limited confidence in current defenses.
A vendor-commissioned SecureIQLab draft reports F5 AI Guardrails blocked 98.36% of 19,679 test payloads across 10 AI security categories, with a 17% false-positive rate. The draft says its findings remain subject to validation and revision.
Hoxhunt reports that phishing samples with indicators of AI assistance rose from 4% in November 2025 to 56% in December, then fell to 40% in January 2026. The report also examines phishing tactics and training outcomes.
WatchGuard’s H1 2026 report says 95.72% of Q2 endpoint malware detections appeared on one machine, while network attacks targeted old flaws. It also found 95% of detected network malware used TLS, which only 20% of Fireboxes inspect.
Apiiro found secrets in 1,122 public copies of Claude Code’s settings.local.json. Researchers did not test whether most remain valid; one notified company confirmed its key was live. Git history and forks can retain the files.
Censys’ 2026 report estimates about 134,000 internet-exposed ICS hosts globally, up 4% since 2024. It identifies mobile-network hosting, expanded attack surfaces, weak configurations, and network concentration as factors sustaining exposure.
CERT-AGID analyzed 217 malicious campaigns affecting Italy in one week (187 targeting Italy, 30 generic) and shared 1,097 IoCs with accredited entities. Unpaid-fine phishing was prominent; AgentTesla and Remcos were among the malware reported.
CYFIRMA’s 2025–2026 assessment finds government was the most-reported victim sector in Australia and New Zealand, with incidents often involving edge systems, credentials and suppliers. It combines incident data, monitoring and campaign analysis, while cautioning that a
Recorded Future's October 2026 review identifies 82 high-impact CVEs, including 69 actively exploited or operationally weaponized. It highlights PHP2Shell exploitation, ransomware activity, and vulnerability trends.
EI-ISAC reported limited cyber, physical and disruptive activity on Nov. 3, 2026, with no major disruptions. Election offices saw unsuccessful probing and password guessing, plus website flooding; adding a CAPTCHA reduced traffic.
Censys measured a 53% decline in Internet-exposed Danfoss StoreView and Copeland/Dixell XWEB controllers from Oct. 9 to Nov. 5, amid reported attacks in France and other European countries. The data does not establish why devices went offline.
Microsoft detected 7.3 billion email phishing threats in Q3 2026. QR-code phishing fell 67% from Q2, while CAPTCHA-gated attacks rebounded. The report details impersonation trends, campaigns, defenses, detections, and hunting queries.
A Northeastern University study found Vizio and Samsung TVs appeared to send ACR-related traffic after opt-out. Researchers also found unencrypted VIDAA detection results accessible without login, though encryption limited verification of other traffic.
Infoblox analyzes three casino-site ecosystems: illegal Chinese-language gambling tied to laundering, scam gambling sites, and decoys hiding PeckBirdy C2. It reports infrastructure trends, detection gaps and network signals defenders can use to triage possible exposure.
Infoblox observed about 50,400 re-registered dropcatch domains daily across gTLDs in the first half of 2026. These domains can inherit reputation and lingering connections that threat actors may abuse.
DarkOwl analyzed NoName057(16)’s English Telegram channel from October 2025 to September 2026, finding 666 DDoS claim posts and growing partner amplification. The group’s victim and intrusion claims remain unverified.
ReliaQuest says manufacturers were named in 563 ransomware attacks in the latest quarter, up 71% quarter over quarter. It also reports more exploited vulnerabilities and identity attacks, and recommends stronger exposure, identity and response controls.
The NCSC’s 2026 annual review assesses that AI-powered malware is highly likely to be used in ransomware against UK critical infrastructure by 2028, and warns AI could accelerate existing attack methods and vulnerability discovery.
CYFIRMA counted 1,178 publicly disclosed ransomware victim claims in October 2026, its highest monthly total, and described shifts in edge access, hypervisor pivots and backup sabotage. Claims were not independently verified.
