# hardening Published articles for hardening. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Kubernetes v1.37: Hardening Container Storage with Bind Mount Options and EmptyDir Permissions DevFeed: [Kubernetes v1.37: Hardening Container Storage with Bind Mount Options and EmptyDir Permissions]() Original publisher: [Read original article]() Author: Nispriha Jagan; Neeraj Krishna Gopalakrishna Published: 2026-09-16T18:30:00Z Content type: article Language: en Sources: [Kubernetes Blog]() Topics: [Kubernetes](), [Security](), [mount](), [Filesystems](), [chmod](), [Unix]() Tags: [chmod](), [container](), [containers](), [filesystem](), [hardening](), [kubernetes](), [linux](), [mount](), [permissions](), [storage](), [volume](), [volumes]() ### AI overview Kubernetes v1.37 adds bind mount options and emptyDir permission modes to strengthen storage security. The article explains how noexec, nosuid, nodev, Unix permissions, and the sticky bit can help enforce security policies on writable volumes. ### Source excerpt Kubernetes v1.37 brings important storage security features: emptyDir permission modes and bind mount options. They help application programmers and security professionals implement rigorous security policies, for example, prohibiting deletion of files across containers or execution of arbitrary binaries from writable volumes, directly in Kubernetes without any complicated circumvention. Linux storage and permission fundamentals Before diving into the new Kubernetes features, let us briefly review the low-level Linux security mechanisms that make them possible. Bind mount flags When Linux mounts or remounts a directory, Virtual File System (VFS) flags control what actions are permitted on that filesystem: noexec: Do not permit direct execution of any binaries on the mounted filesystem. nosuid: Do not allow set-user-identifier or set-group-identifier bits to take effect. nodev: Do not interpret character or block special devices on the file system. Directory permissions and the sticky bit Standard Unix permissions regulate access across three scopes: Owner, Group, and Others (e.g., 0755 or 0777). Beyond standard read, write, and execute bits, Linux supports the sticky bit (as in mode 01777). When applied to a directory, the sticky bit ensures that a file inside that directory can only be deleted or renamed by the file's owner or root. This is essential for shared writable directories like /tmp. Motivation for the improvements Why does Kubernetes need bind mount options and emptyDir permissions? The primary goal of these features is to increase the security of Kubernetes workloads by allowing security-related bind mount options on volume mounts. By default, volumes are bind-mounted into containers by the container runtime and kubelet without noexec, nosuid, or nodev flags. This default can undermine security. For example, with noexec missing, a compromised process can use any writable volume (emptyDir, PersistentVolume, etc.) to download, chmod +x, and execute arbitra ## Cisco and the DISA STIG: Turning Zero Trust Policy into Repeatable Practice - Part 2: Cisco SNA DevFeed: [Cisco and the DISA STIG: Turning Zero Trust Policy into Repeatable Practice - Part 2: Cisco SNA]() Original publisher: [Read original article]() Author: Norman St. Laurent Published: 2026-09-15T13:13:53Z Content type: article Language: en Sources: [Cisco Blogs]() Topics: [Cybersecurity](), [Cisco](), [telemetry](), [Zero Trust](), [vulnerability](), [Network]() Tags: [benchmark](), [cisco](), [cisco-secure-network-analytics-sna](), [cybersecurity](), [department-of-defense-dod](), [government](), [hardening](), [industries](), [nist](), [public-sector](), [stig](), [visibility]() ### AI overview The article explains how the DISA Security Technical Implementation Guide for Cisco Secure Network Analytics turns Zero Trust policy into testable configuration requirements. The STIG provides a shared hardening baseline for the platform and its management functions, with 31 requirements derived from NIST SP 800-53 and related requirements. ### Source excerpt Discover how the new DISA STIG for Cisco Secure Network Analytics helps defense organizations securely configure and harden their analytics platform, ensuring trusted network visibility for Zero Trust operations. ## Announcing Chainguard container images for Go 1.27 DevFeed: [Announcing Chainguard container images for Go 1.27]() Original publisher: [Read original article]() Published: 2026-08-31T00:00:00Z Content type: news Language: en Sources: [Chainguard: Unchained]() Topics: [container images](), [Go Language](), [Cryptography](), [Deployment](), [Security, Privacy and Abuse Prevention](), [Post-Quantum]() Tags: [chainguard](), [container](), [container-images](), [cryptography](), [deployment](), [fedramp](), [fedramp-ato](), [fips](), [go](), [hardening](), [images](), [post-quantum](), [release]() ### AI overview Chainguard announces three Go 1.27 container images: go, go-fips, and go-openssl-fips. The options differ in their FIPS-validated cryptography, including whether cryptography is built in or selected at deployment time. ### Source excerpt Chainguard simplifies Go 1.27 with three images and a new option to choose FIPS-validated cryptography at deployment without recompiling. ## Tapjacking Protection: Rejecting Android Touches Behind an Overlay DevFeed: [Tapjacking Protection: Rejecting Android Touches Behind an Overlay]() Original publisher: [Read original article]() Author: Shai Almog Published: 2026-08-25T00:00:00Z Content type: release Language: en Sources: [CodeName One]() Topics: [Android](), [Security]() Tags: [android](), [android-12](), [false-positive](), [hardening](), [protection](), [report]() ### AI overview Codename One adds Android tapjacking and screen-overlay protection. The feature can report or block gestures that begin behind another application's overlay and can ask Android 12 or newer to prevent overlays on sensitive screens. ### Source excerpt Codename One can now report or block Android gestures that begin behind another application's overlay, and can ask Android 12 or newer to hide overlay windows on sensitive screens. ## App Hardening: One Obfuscation Pipeline Across Every Port DevFeed: [App Hardening: One Obfuscation Pipeline Across Every Port]() Original publisher: [Read original article]() Author: Shai Almog Published: 2026-08-15T00:00:00Z Content type: article Language: en Sources: [CodeName One]() Topics: [obfuscation](), [cross-platform](), [Encryption](), [Security](), [Reverse Engineering](), [App]() Tags: [android](), [app](), [cross-platform](), [encryption](), [hardening](), [ios](), [javascript](), [native](), [obfuscation](), [platform](), [r8](), [reverse-engineering](), [security]() ### AI overview Codename One App Hardening applies one pre-build protection pipeline to a merged cross-platform application before it is split into Android, iOS, JavaScript, Windows, Linux, and desktop builds. It supports renaming, eligible string encryption, selected control-flow transformations, verification, and crash-report retracing, while acknowledging that reverse engineering cannot be made impossible. ### Source excerpt Codename One App Hardening applies class and member renaming, string encryption, and platform-safe control-flow transforms before one application becomes Android, iOS, JavaScript, or native desktop output. ## Where Apollo MCP Server Stands on the OWASP MCP Top 10 DevFeed: [Where Apollo MCP Server Stands on the OWASP MCP Top 10]() Original publisher: [Read original article]() Author: Camille Lawrence Published: 2026-08-14T17:11:10Z Content type: article Language: en Sources: [Apollo Blog]() Topics: [Model Context Protocol](), [MCP Server](), [audit](), [Shared Responsibility Model](), [configuration](), [GraphQL](), [Rust](), [Self-hosted]() Tags: [announcement](), [apollo](), [apollo-mcp-server](), [audit](), [configuration](), [graphql](), [hardening](), [mcp](), [mcp-server](), [rust](), [self-hosted](), [shared-responsibility]() ### AI overview This article evaluates Apollo MCP Server against all ten risks in the latest reviewed OWASP MCP Top 10 revision. It finds that the server's architecture directly closes one risk because typed GraphQL operations are parsed and type-checked without shell execution. The remaining risks require configuration, monitoring, and team-owned security decisions. ### Source excerpt Architecture alone closes 1 of the ten risks in OWASP's MCP Top 10. See how Apollo MCP Server scores on the rest, and which configuration choices remain yours. ## How Mobile App Security from Guardsquare Addresses Gaps in Framework Compliance DevFeed: [How Mobile App Security from Guardsquare Addresses Gaps in Framework Compliance]() Original publisher: [Read original article]() Author: Guest post by Dr. Edward Amoroso, CEO, TAG Infosphere Inc. and former AT&T Chief Security Officer Published: 2026-08-11T14:00:31Z Content type: opinion Language: en Sources: [Guardsquare Blog]() Topics: [Mobile Security](), [Application Security](), [Security](), [Cybersecurity](), [Artificial Intelligence]() Tags: [application-security](), [compliance](), [credentials](), [cybersecurity](), [frameworks](), [governance](), [hardening](), [mobile](), [protection](), [regulatory](), [report](), [security](), [standards]() ### AI overview A featured guest post examines how mobile application security, including Guardsquare's hardening, runtime protection, and anti-tampering capabilities, can help enterprises address gaps in compliance frameworks. It also argues that frameworks such as NIST CSF 2.0 should more explicitly account for mobile app risk. ### Source excerpt Former AT&T Chief Security Officer and TAG Infosphere founder Dr. Edward Amoroso shares his perspective on the state of mobile application security in a featured guest post for Guardsquare. Enterprise compliance is evolving as organizations face mounting regulatory pressure and more capable threat actors. Regulators now expect alignment to frameworks such as National Institute of Standards and Technology Cybersecurity Framework (CSF) 2.0 and sector-specific mandates. Yet, while governance has matured around cloud, endpoint, and networks, mobile app risk remains underrepresented in compliance frameworks and control processes. ## Vaultwarden: gestor de contraseñas autoalojado DevFeed: [Vaultwarden: gestor de contraseñas autoalojado]() Original publisher: [Read original article]() Author: Antonio Pérez Published: 2026-08-03T00:00:00Z Content type: tutorial Language: es Sources: [tengoping.com]() Topics: [Docker](), [Docker Compose](), [TLS (Transport Layer Security)](), [nginx](), [proxy](), [backups]() Tags: [backups](), [bitwarden](), [compose](), [docker](), [docker-compose](), [hardening](), [homelab](), [https](), [nginx](), [proxy](), [tls](), [vaultwarden]() ### AI overview A Spanish-language tutorial explains how to self-host Vaultwarden, an unofficial Rust server compatible with Bitwarden clients. It covers Docker Compose deployment, HTTPS behind an Nginx reverse proxy, admin-token hardening, and backups, while noting the trade-off between community support and the official Bitwarden server. ### Source excerpt Instala Vaultwarden con Docker: servidor ligero compatible con los clientes de Bitwarden, con HTTPS, hardening del admin token y backups. ## How Harness Secures Autonomous Worker Agents in Production DevFeed: [How Harness Secures Autonomous Worker Agents in Production]() Original publisher: [Read original article]() Author: Shivansh Srivastava Shubham Jindal Published: 2026-07-13T00:00:00Z Content type: article Language: en Sources: [Harness Blog]() Topics: [Security](), [Architecture & Design](), [Processes](), [Network](), [Exploit]() Tags: [agent](), [exploit](), [hardening](), [kernel](), [network](), [security]() ### AI overview Harness describes four security layers for Autonomous Worker Agents: hardened images, process isolation, secret isolation through a broker, and network controls through a proxy. The company says it tested the design against a real CVSS-9.0 breach chain, reducing access from 709 live secrets to 33 variables and zero usable credentials. ### Source excerpt How Harness isolates Autonomous Worker Agents with four independent security layers -- image hardening, process isolation, secret isolation, and network controls | Blog ## Chainguard Agent Skills is now open to everyone, with a private registry to manage your internal skills DevFeed: [Chainguard Agent Skills is now open to everyone, with a private registry to manage your internal skills]() Original publisher: [Read original article]() Published: 2026-06-17T00:00:00Z Content type: article Language: en Sources: [Chainguard: Unchained]() Topics: [Agent Skills](), [chainguard](), [Security](), [audit trail](), [Open Source]() Tags: [agent-skills](), [ai-agent-skills](), [audit-trail](), [chainguard](), [chainguard-agent-skills](), [chainguard-ai-agents](), [chainguard-skills](), [cursor](), [hardening](), [private-registry](), [registry](), [security]() ### AI overview Chainguard Agent Skills is now available to all users with more than 1,000 hardened community skills, a private registry for internal skills, and a closed beta for hardening first-party skills. The service is intended to help organizations securely standardize, distribute, and manage AI agent skills. ### Source excerpt Chainguard Agent Skills now offers 1,000+ hardened community skills, private registries, and beta hardening for first-party AI agent skills. ## The expanding threat landscape: Chainguard now scans source code for traditional malware and "greyware" DevFeed: [The expanding threat landscape: Chainguard now scans source code for traditional malware and "greyware"]() Original publisher: [Read original article]() Published: 2026-06-09T00:00:00Z Content type: article Language: en Sources: [Chainguard: Unchained]() Topics: [chainguard](), [Malware](), [Code](), [Open Source](), [npm](), [npm packages](), [Security]() Tags: [chainguard](), [chainguard-libraries](), [ecosystems](), [greyware](), [greyware-scanner](), [hardening](), [javascript](), [malware](), [malware-prevention](), [malware-scanner](), [npm](), [npm-packages](), [open-source](), [packages](), [scanner](), [security](), [source](), [supply-chain-attacks]() ### AI overview Chainguard has introduced a source code scanner that detects traditional malware and "greyware," harmful packages that may perform actions such as credential theft, command interception, API key harvesting, or persistent remote access. The scanner currently protects npm packages, scans more than 100,000 packages daily, and has blocked more than 52,000 packages identified as malware or greyware. ### Source excerpt Chainguard's new scanner blocks malware and 'greyware' before it reaches developers, protecting 100,000+ packages daily across open source ecosystems. ## wasmCloud 2.2.0: TLS for WebAssembly, customizable HTTP egress DevFeed: [wasmCloud 2.2.0: TLS for WebAssembly, customizable HTTP egress]() Original publisher: [Read original article]() Author: Eric Gregory Published: 2026-05-21T00:00:00Z Content type: release Language: en Sources: [wasmCloud Blog]() Topics: [WebAssembly](), [HTTP](), [TLS (Transport Layer Security)]() Tags: [config](), [handler](), [hardening](), [http](), [middleware](), [release](), [rpc](), [tls](), [wasi](), [webassembly]() ### AI overview wasmCloud 2.2.0 adds native TLS for WebAssembly components through WASI Preview 3, a pluggable HTTP egress handler, an expanded wash config command, and runtime-operator hardening. The release also supports customizable TLS providers and embedder-controlled outbound request handling. ### Source excerpt wasmCloud 2.2.0 brings WASI Preview 3 TLS to WebAssembly components, a pluggable OutgoingHandler trait for HTTP egress, an expanded wash config command, and runtime-operator hardening. ## Chainguard brings first-party RHEL 9 and RHEL 10 RPM support to Chainguard OS, joins FINOS DevFeed: [Chainguard brings first-party RHEL 9 and RHEL 10 RPM support to Chainguard OS, joins FINOS]() Original publisher: [Read original article]() Published: 2026-05-11T00:00:00Z Content type: news Language: en Sources: [Chainguard: Unchained]() Topics: [chainguard](), [chainguard os](), [chainguard containers](), [Security](), [Critical Infrastructure](), [Package Management](), [Open Source](), [Cloud Native Ecosystem](), [Resilience](), [APK](), [Linux]() Tags: [chainguard](), [chainguard-containers](), [chainguard-finserv](), [chainguard-os](), [cloud](), [cloud-native](), [compatibility](), [compliance](), [container](), [container-images](), [containers](), [critical-infrastructure](), [cve](), [dependencies](), [hardening](), [images](), [linux](), [make](), [mythos](), [open](), [open-source](), [project-glasswing](), [rhel](), [rhel-10](), [rhel-9]() ### AI overview Chainguard announces first-party RHEL 9 and RHEL 10 RPM compatibility for packages in Chainguard Containers built on Chainguard OS. The company also announces that it is joining FINOS to support open source collaboration in financial infrastructure. ### Source excerpt Chainguard adds first-party RHEL 9/10 RPM compatibility and joins FINOS, helping financial institutions modernize securely for the AI-driven threat era. ## Protocol Cluster Updates: May 2026 DevFeed: [Protocol Cluster Updates: May 2026]() Original publisher: [Read original article]() Author: Will Corcoran; Kev Wedderburn; Fredrik Published: 2026-05-11T00:00:00Z Content type: article Language: en Sources: [Ethereum Foundation Blog]() Topics: [Ethereum](), [client](), [Post-Quantum]() Tags: [ethereum](), [hardening](), [interop](), [leadership](), [norway](), [research-development](), [team](), [updates](), [upgrade]() ### AI overview This May 2026 update reports on an Ethereum core developer interop in Svalbard, where teams worked on hardening and preparing for the Glamsterdam upgrade. It highlights progress on the 200M gas limit floor, ePBS, EIP-8037, and Hegotá groundwork, and announces a transition in Ethereum Foundation Protocol cluster leadership. ### Source excerpt A semi-regular gathering of Ethereum core devs from various client teams, or interop, recently took place in Svalbard, Norway. Over the week-long event, teams focused on hardening and preparation for the next upgrade, Glamsterdam. Several important milestones came out of the week, including: 200M gas limit floor established:... ## An exploitable integer overflow in Lix (CVE-2026-44028) DevFeed: [An exploitable integer overflow in Lix (CVE-2026-44028)]() Original publisher: [Read original article]() Published: 2026-05-05T00:00:00Z Content type: release Language: en Sources: [News on Lix]() Topics: [Security](), [Vulnerabilities](), [Exploit](), [Nix](), [Incident response]() Tags: [cve](), [exploit](), [hardening](), [incident-response](), [security](), [update]() ### AI overview Lix has an integer-overflow vulnerability assigned CVE-2026-44028. The issue can enable an out-of-bounds write through a specially crafted NAR archive, with exploitability depending on build type and ASLR defenses. Fixed versions are Lix 2.93.4, 2.94.2, and 2.95.2, and users are advised to update. ### Source excerpt Security researchers have found a security issue in Lix. This issue has been assigned CVE-2026-44028. Important note : The issues are different between Lix and CppNix but it seems there was confusion in MITRE who emitted the CVE and copied the wrong information which should have gone into the CppNix CVE, we are trying to update the CVE metadata. ## SSH Port 22: Custom Ports, Port Forwarding Security, and Production SSH Hardening DevFeed: [SSH Port 22: Custom Ports, Port Forwarding Security, and Production SSH Hardening]() Original publisher: [Read original article]() Author: info@goteleport.com (Matt Keib) Published: 2026-05-01T00:00:00Z Content type: tutorial Language: en Sources: [Teleport]() Topics: [Security](), [OpenSSH](), [ssh](), [Authentication](), [rate-limiting](), [Authorization](), [configuration]() Tags: [configuration](), [hardening](), [port](), [rate-limiting](), [security](), [server](), [ssh](), [tcp](), [tcp-handshake]() ### AI overview This article explains that SSH port 22 is a historical TCP entry point with no inherent cryptographic or authentication significance. It covers port forwarding risks, cryptographic and authentication configuration, rate limiting, and production SSH hardening. ### Source excerpt Learn about SSH port 22, its limitations, port forwarding security, and how to harden production SSH. ## Supply chain attacks expose weaknesses in open-source software trust DevFeed: [Supply chain attacks expose weaknesses in open-source software trust]() Original publisher: [Read original article]() Published: 2026-04-09T00:00:00Z Content type: opinion Language: en Sources: [Chainguard: Unchained]() Topics: [Open Source](), [vulnerability](), [Exploit](), [npm]() Tags: [ai-coding](), [attacks](), [chainguard](), [code](), [dependency](), [hardening](), [npm](), [open-source](), [open-source-security](), [oss](), [pypi](), [scanner](), [secrets](), [software-supply-chain-security](), [solarwinds](), [supply-chain](), [supply-chain-attacks](), [trust](), [vibe-coding](), [vulnerability](), [vulnerability-scanner]() ### AI overview This opinion article argues that a series of open-source supply chain attacks exposed a broken trust model. It discusses attacks involving npm and PyPI packages, stolen credentials, malicious contributors, and the limits of vulnerability scanning and hardening. ### Source excerpt Five supply chain attacks in 12 days exposed a broken trust model. Learn why scanning and hardening fail, and why trusting the source is the only fix. ## Introducing Chainguard Agent Skills: Because your AI agent shouldn't trust strangers DevFeed: [Introducing Chainguard Agent Skills: Because your AI agent shouldn't trust strangers]() Original publisher: [Read original article]() Published: 2026-03-17T00:00:00Z Content type: article Language: en Sources: [Chainguard: Unchained]() Topics: [chainguard agent skills](), [Agent Skills](), [supply-chain-security](), [chainguard](), [Security](), [Vulnerabilities](), [Malware]() Tags: [ai-agents](), [anthropic](), [chainguard](), [chainguard-agent-skills](), [chainguard-ai-tools](), [chatgpt](), [claude](), [common-vulnerabilities-and-exposures](), [hardened-ai-agent-skills](), [hardening](), [malware](), [security](), [supply-chain-security]() ### AI overview Chainguard Agent Skills is introduced as a continuously maintained catalog of hardened AI agent skills. The catalog applies review, permission scoping, integrity verification, continuous hardening, and audit trails to reduce the supply-chain risks of third-party skills used with platforms such as Claude Code and OpenClaw. ### Source excerpt Chainguard Agent Skills is a continuously maintained catalog of hardened AI agent skills. ## Auditoría de seguridad con Lynis en Linux DevFeed: [Auditoría de seguridad con Lynis en Linux]() Original publisher: [Read original article]() Author: Antonio Pérez Published: 2026-02-07T00:00:00Z Content type: tutorial Language: es Sources: [tengoping.com]() Topics: [Linux]() Tags: [hardening](), [linux]() ### AI overview A Spanish tutorial on using Lynis to audit Linux server security, interpret its hardening score and findings, and apply recommendations. ### Source excerpt Cómo usar Lynis para auditar la seguridad de servidores Linux, interpretar resultados y aplicar las recomendaciones. ## Hardening básico de servidores Linux DevFeed: [Hardening básico de servidores Linux]() Original publisher: [Read original article]() Author: Antonio Pérez Published: 2026-01-22T00:00:00Z Content type: tutorial Language: es Sources: [tengoping.com]() Topics: [Linux](), [Firewall](), [ssh](), [Fail2ban](), [OpenSSH](), [Kernel]() Tags: [fail2ban](), [firewall](), [hardening](), [hash](), [kernel](), [linux](), [ssh]() ### AI overview This Spanish tutorial presents a basic hardening checklist for Linux servers and explains why the measures should be applied in a specific order. It covers system updates, SSH, firewall configuration, user policies, sudo restrictions, kernel parameters, unnecessary services, Fail2ban, auditing, and final verification. ### Source excerpt Checklist de hardening en el orden correcto: SSH, firewall, usuarios, kernel, AIDE, chrony y verificación final con Lynis. ## Why Trusted Software Supply Chains Matter More Than Zero-CVE Container Claims DevFeed: [Why Trusted Software Supply Chains Matter More Than Zero-CVE Container Claims]() Original publisher: [Read original article]() Published: 2026-01-15T00:00:00Z Content type: opinion Language: en Sources: [Chainguard: Unchained]() Topics: [Security](), [container-security](), [container images](), [Docker Hub](), [Containers](), [Docker](), [Software]() Tags: [ceo](), [chainguard](), [chainguard-containers](), [chainguard-images](), [chainguard-libraries](), [container-images](), [docker](), [docker-hub](), [docker-images](), [echo-security](), [hardened-images](), [hardening](), [minimus](), [rapidfort](), [secure-by-default](), [security](), [software-supply-chain-security](), [wiz-images](), [zero-cves]() ### AI overview Chainguard CEO Dan Lorenc argues that container security depends on trusting the origins and build processes of software, rather than relying primarily on post-hoc image hardening or zero-CVE claims. ### Source excerpt Chainguard CEO Dan Lorenc explains why real security comes from trusted, from-source software supply chains, not post-hoc hardening or zero-CVE promises. ## Cómo configurar un servidor SSH seguro en Linux DevFeed: [Cómo configurar un servidor SSH seguro en Linux]() Original publisher: [Read original article]() Author: Antonio Pérez Published: 2026-01-01T00:00:00Z Content type: tutorial Language: es Sources: [tengoping.com]() Topics: [OpenSSH](), [Linux](), [Fail2ban]() Tags: [fail2ban](), [firewall](), [hardening](), [linux](), [openssh]() ### AI overview A step-by-step guide to securing an OpenSSH server on Linux. It covers layered protections including non-standard ports, public-key authentication, hardened ciphers, two-factor authentication, and fail2ban, along with configuration and verification guidance. ### Source excerpt Guía paso a paso para configurar y securizar un servidor SSH en cualquier servidor Linux, incluyendo autenticación por clave, fail2ban y mejores prácticas. ## Continuously hardening ChatGPT Atlas against prompt injection DevFeed: [Continuously hardening ChatGPT Atlas against prompt injection]() Original publisher: [Read original article]() Published: 2025-12-22T00:00:00Z Content type: article Language: en Sources: [OpenAI News]() Topics: [prompt injection](), [Security](), [ai security](), [ChatGPT](), [OpenAI](), [browser](), [Adversarial attacks](), [Reinforcement learning]() Tags: [agent](), [agentic](), [ai](), [ai-security](), [atlas](), [attacks](), [chatgpt](), [hardening](), [prompt-injection](), [reinforcement-learning](), [security](), [techniques](), [tool](), [web]() ### AI overview OpenAI describes how it is continuously hardening ChatGPT Atlas against prompt injection. Automated red teaming powered by reinforcement learning helps discover novel attacks and exploits so the browser agent's defenses can be improved before attacks spread in the wild. ### Source excerpt OpenAI is strengthening ChatGPT Atlas against prompt injection attacks using automated red teaming trained with reinforcement learning. This proactive discover-and-patch loop helps identify novel exploits early and harden the browser agent's defenses as AI becomes more agentic. ## ECS on EC2: Covering Gaps in IMDS Hardening DevFeed: [ECS on EC2: Covering Gaps in IMDS Hardening]() Original publisher: [Read original article]() Published: 2025-10-02T18:18:59Z Content type: tutorial Language: en Sources: [Latacora]() Topics: [Amazon Elastic Container Service](), [Amazon EC2](), [Security](), [Amazon Web Services](), [Containers](), [Credential theft]() Tags: [aws](), [containers](), [credential-theft](), [credentials](), [ec2](), [ecs](), [hardening](), [security](), [sensitive-data]() ### AI overview This article examines security gaps in Amazon ECS workloads running on EC2, focusing on task isolation and restricting access to the EC2 Instance Metadata Service. It discusses how weak isolation can expose credentials and sensitive data and outlines the need for more comprehensive hardening guidance. ### Source excerpt Introduction # AWS ECS is a widely-adopted service across industries. To illustrate the scale and ubiquity of this service, over 2.4 billion Amazon Elastic Container Service tasks are launched every week (source) and over 65% of all new AWS containers customers use Amazon ECS (source). There are two primary launch types for ECS: Fargate and EC2. The choice between them depends on factors like cost, performance, operational overhead, and the variability of your workload. [Next page]()