Hello, stay ahead with CISO Brief 🚀

🛡️ Researchers at Parameter disclosed a flaw in ChatGPT shared links that allowed attacker-supplied code to execute inside the signed-in session of anyone who opened the link. Reported to OpenAI on October 18 and fixed eight days later as a P1 severity, the attack required only opening a shared conversation and, in some variants, clicking the Preview button. The exploit chained a Preview-generated iframe with a vulnerable MathLive listener to inject HTML into the main chat page, letting the attacker call ChatGPT's signed-in APIs and access private conversations. OpenAI patched the issue across web frontends by rejecting the malicious messages; users need not take action. Parameter published details on October 28 and received a $5,000 bounty.

📊 Cloudflare reported it mitigated 18.2 million network-layer (L3) DDoS attacks and 14.4 trillion HTTP DDoS requests in Q3 2026, averaging one attack every 0.44 seconds. The firm ties the surge to DDoS-for-hire booter services and notes a 208% rise in short-burst, low-bandwidth attacks under 10 minutes and 1 Gbps. While hyper-volumetric attacks over 1 Tbps declined 69% from Q2—likely due to law enforcement actions—new variants such as Dysphoria and Evooo1Bot are emerging. Key targets include media, IT, gambling, software, and telecoms across the U.S., China, Indonesia, the U.K., and Türkiye.

🛡️ Buyers in 2026 face consolidated WAAP choices as platforms aim to replace separate WAF, API, bot, DDoS, and client-side controls with a single service. New AMTSO-certified test data from SecureIQLab v5.0 highlights large variance in overall protection and false-positive rates across vendors, making efficacy and tuning effort critical selection criteria. The article recommends probing API protocol coverage, time-to-tune on real traffic, and deployment/licensing scalability during proof-of-value trials.

🔔 Amazon Bedrock now supports OpenAI GPT-5.6 (Terra and Luna) on the bedrock-runtime endpoint in AWS GovCloud (US), and adds cross-Region inference to increase throughput and redundancy. The update enables the Responses, Converse, and Chat Completions APIs for these models and integrates model usage with existing Bedrock account-level controls. Invocation logs and CloudWatch metrics capture usage, latency, and errors, and costs are itemized in Cost Explorer and the Cost and Usage Report. Customers can review GPT-5.6 model cards in the Amazon Bedrock User Guide to get started.

🔎 Kiro web search is now available in AWS GovCloud (US), giving agents real-time access to current internet information so responses can include up-to-date search results and page content. The agent automatically uses web search for queries involving recent developments or topics outside the model's training data, and answers that reference web results include citations. Enterprise administrators can disable web search and web fetch for their organization via the Kiro console under Settings, and teams can remove web tools from individual custom agents or block them using capability-based permission rules. For further details, consult the AWS GovCloud (US) documentation, the monitoring and tracking guide, or contact your AWS account team.

📧 Amazon Connect Customer now automatically saves in-progress emails as drafts so agents can resume work after interruptions without losing content. Drafts restore the message body, recipients, subject line, and attachment references when agents return. Saved draft data is also stored in customer Amazon S3 buckets to enable integrations with other applications and workflows.

🔒 Daybreak Red from OpenAI is now available to eligible customers on Amazon Bedrock in AWS GovCloud (US‑East), providing an AI model tailored for authorized vulnerability research, exploit reproduction, and mitigation development. Built as part of the Daybreak cyber defense initiative, Daybreak Red: GPT‑5.6 Cyber includes stronger identity verification, monitoring, and access controls, and runs on Bedrock’s next‑generation inference engine with zero‑operator access at the chip. Inference data is not used to train models and Daybreak Red data need not be shared with OpenAI. Access requires enrollment through OpenAI and coordination with your AWS account team.

🔒 Amazon CloudFront can now forward two optional headers to origins: CloudFront-Viewer-ISP, which reports the viewer's internet service provider, and CloudFront-Viewer-TLS-Details, which reports the cryptographic details of the viewer TLS connection, including post-quantum indicators. These headers enable traffic and performance breakdowns by network and measurement of post-quantum TLS adoption. To use them, add the desired header(s) to an origin request policy and attach that policy to your distribution.

🔐 The identity landscape now spans humans, service accounts, machine identities, and AI agents, each representing an access decision that must be governed. Boards and regulators expect security leaders to demonstrate who or what has access to critical systems and how access is controlled. Continuous identity security replaces periodic reviews with ongoing discovery, lifecycle management, and evidence collection to maintain up-to-date oversight and meet accountability demands.

🔔 Amazon RDS for Oracle now allows you to disable automatic minor version upgrades for DB instances using the JVM, Spatial, Locator, or OLAP options. This change applies to new and existing instances—previously these options required automatic upgrades. You should still apply minor upgrades regularly for security fixes, or use AWS Organizations upgrade rollout policy to manage upgrade order while keeping automatic upgrades enabled.

🔒 Amazon MSK now supports TLS 1.3 for MSK Provisioned clusters via named encryption policies. Customers can opt in by selecting a security policy that defines permitted TLS versions and cipher suites, with no change to existing clusters until a policy is applied. Policies can permit both TLS 1.2 and 1.3 to enable gradual client upgrades or enforce stricter TLS 1.3-only requirements for compliance.

📘 Cybercriminals exploit ebook demand through fake downloads, malicious files, and phishing schemes targeting readers. Pirate and search-spam sites often deliver malware disguised as books or steal credentials via malicious archives and executables. Phishing campaigns impersonate legitimate book communities and stores to harvest payments and personal data, while fake library mirrors can leak millions of user records. Protect devices by using official stores, keeping software updated, avoiding dubious sites, and backing up your library.

🔒 Run a Google Security Checkup to see who has access to your Gmail inbox in about two minutes. Review signed-in devices, remove outdated third-party app access, and confirm your recovery information to secure your account. These three steps help reduce phishing risk and prevent unauthorized access. For additional guidance, consult the Gmail Help Center on protecting your account.

🔒 Today Google Cloud announces that spend caps on Cloud Budgets are generally available, enabling cost controls that avoid resource deletion. The release adds three key enhancements: a spend caps API for programmatic enforcement, live cost reports for near-real-time visibility into triggers, and automated monthly resets to remove the need for manual unlocking. These updates aim to make guardrails scalable and operationally efficient for organizations managing many projects.

🛡️ Automatic SDK integration for AWS WAF injects the AWS WAF JavaScript SDK into HTML served by Amazon CloudFront when enabled, requiring no changes to application code, build processes, or deployments. The client-side signals provided by the SDK support three AWS WAF managed rule groups: Bot Control, Fraud Control ATP, and Fraud Control ACFP. The feature removes the need to add and maintain script tags across many applications or pages you don't control. It can be configured via the Console, CLI, SDK, CDK, or CloudFormation and is free to use, with guidance available in the AWS WAF Developer Guide.

🛡️ The Ransomware Assessment and Incident Defense Framework for AI Systems (RAID-AI) is now freely available online to help security leaders identify weaknesses that could expose AI systems to ransomware. Developed by Behnaz Karimi and Yuvaraj Govindarajulu of OWASP AIBOM, the hands-on framework asks teams to self-assess controls across general cybersecurity and AI-specific areas. RAID-AI maps controls to NIST CSF 2.0 and CISA CPG 2.0 and supports attack mapping to MITRE ATT&CK, ATLAS and NIST Adversarial, presenting results in dashboards and downloadable reports.

🔒 TeamViewer patched a high-severity flaw (CVE-2026-111585, 8.8) in version 15.83 that allowed a remote host to crash or potentially execute code on the viewer after a remote-control session is started. The bug affected Full Client on Windows, Linux, macOS, Windows Portable, and Android Remote Control app in versions prior to 15.83. Researcher Omri Dar of Remedio published technical details showing a screen-size calculation overflow that enables memory corruption; TeamViewer recommends updating as the only mitigation.

🔍 Fortinet and UC Berkeley’s Center for Long-Term Cybersecurity examined tabletop exercises exploring how AI-enabled cybercrime stresses decision-making across detection, escalation, and response. The collaboration highlighted recurring governance gaps—authority, information sharing, and timing—that slow action even when threats are detected. The exercises showed the value of pre-defined decision roles, tested coordination with partners, and emphasized pairing faster detection with clear human accountability.