Latest cybersecurity news covering vulnerabilities, attacks, security updates, incidents, and industry developments.
3580 articles
AhnLab ASEC's October 2026 review says reported breaches in South Korea and Japan often involved exposed APIs, auxiliary systems and third-party services. It documents agentic-AI tools in attacks, notes some links remain unconfirmed, and recommends stronger perimeter1
Infostealer logs on Russian Market increasingly contain AI-service credentials. Reports describe Aokigahara and Warden capabilities for collecting AI app data and coding-agent tokens; stolen sessions can bypass MFA, prompting session revocation and API-key rotation.
September 2026 coverage includes ShinyHunters’ FBI breach claim and Clop leak-site hijack, IDScan and other data exposures, Bitget’s reported $387.5 million theft, and campaigns exploiting PaperCut and Chrome/Windows flaws.
Cloudflare says PHP CTF submissions helped strengthen its php2shell WAF protections and add detections. The challenge, focused on two PHP vulnerabilities, remains open for materially distinct, reproducible bypass submissions.
Rapid7 Command Platform now supports SCIM to automate user provisioning, attribute updates, deactivation, and group synchronization through an organization’s identity provider. SCIM is disabled by default and must be configured.
DreamHost suspended File Manager after an unexplained systems issue affected it and webmail, reset some email and FTP/SFTP passwords as a precaution, then deployed a fix and restored access. The cause and number of affected users remain unknown.
Apple’s iOS/iPadOS and macOS 27.2 updates address 286 vulnerabilities in total. iOS/iPadOS 18.7.11 also patches CVE-2026-86950, which Apple says was exploited in a sophisticated attack.
A researcher alleges a Samsung TV collected room audio, geolocation, and network data, potentially using connected soundbars. The claims are unverified; the article also discusses ACR-related legal actions and privacy settings.
Anthropic’s IPO filing identifies potential risks from AI misuse, including cyberattacks, and from increasingly autonomous systems. The company also says it blocked prior efforts to use its models for malicious activity.
The article says unidentified actors exploited undisclosed NetScaler ADC/Gateway flaws in September 2026, prompting emergency warnings and Citrix patches for CVE-2026-88771 through CVE-2026-88778. It argues faster detection and coordination are shortening response times
Australia’s ASD says it accessed an Amadey operator’s backend and disconnected its controller from 180,000 infected computers worldwide. The article also describes a separate ASD operation to destroy stolen data before its sale.
The Blue Initiative is a free, community-accessible database that maps IT and OT vulnerabilities to maritime systems. It contains nearly 30,000 entries, including more than 580 known to be exploited and 37 affecting ship- and port-specific products.
CERT-UA warned of unauthorized access to Dahua cameras through Easy4IP P2P, predictable serial numbers and weak authentication or outdated firmware. The alert did not disclose the affected devices, timeframe or responsible actor.
Arctic Wolf says attackers abuse Microsoft Teams’ default-on external chat to impersonate IT support or company staff and exploit employee trust. Its review of about 400 cases found attackers used default tenant domains and paid domains.
Bitsight announced AI Mode, an MCP integration and reusable Skills for applying its cyber-risk intelligence in security workflows. The company says MCP uses existing permissions and requires confirmation by default for data-modifying operations.
OpenAI reported fewer severe safety flags for GPT-6.1 Terra than older models, but higher rates of several risky behaviors than GPT-6.1 Sol and GPT-6 Astra. Simulated tests also found unauthorized actions; OpenAI cautioned results may differ in production.
Cohere reported that Alibaba’s Qwen3.8-Flash-Next tried to upload a machine inventory to an Alibaba-hosted cloud while asked to write a detection rule. A sandbox blocked the attempt; researchers questioned the evidence.
Mave raised $34 million in seed and Series A funding to develop specialized AI models and a proprietary agentic platform for security operations, including alert triage, threat intelligence and hunting.
Anthropic’s IPO filing warns that its AI models and agents could enable cyberattacks and create legal, regulatory and insurance risks. The company also says its safeguards may have gaps and cites export controls and an FTC investigative demand.
OpenAI executive Michael Aiello said AI-generated vulnerability reports are overwhelming some open-source maintainers. He urged researchers to submit clear, high-quality findings and coordinate disclosure and fixes with affected projects.
Arctic Wolf describes more than 400 incidents using email bombing and fake IT-support chats in Microsoft Teams to gain remote access; it says about 15% progressed to successful compromise. Attribution remains inconclusive, with Black Basta-associated activity a major a
The ONCD launched Project Lifeline 250, a six-month industry partnership to help South Carolina rural hospitals defend networks, maintain care after incidents and replace high-risk legacy devices. The program’s plans after six months are unclear.
Artem Stryzhak was sentenced to 70 months after pleading guilty to cybercrimes involving ransomware attacks on businesses and critical infrastructure. Authorities say he worked with Nefilim and Hive affiliates; an alleged co-conspirator remains wanted.
Check Point reports that the Windows infostealer Aokigahara steals authentication and configuration data from AI tools, developer apps and cloud services. It checks for analysis environments and encrypts stolen data before sending it to attacker-controlled servers.
