OpenAI's launch post (29 September 2026, DevDay San Francisco) introduces dots: always-on agents powered by GPT-6 Astra that run on their own cloud computer and browser, connect to more than 4,000 apps through OpenAI's plugin ecosystem, keep working toward goals 24/7, and are reachable through ChatGPT, Slack and Microsoft Teams (texting coming). With permission a dot can also connect to and use the user's own laptop. When idle, a dot runs 'proactive research' over connected apps with tools OpenAI says are restricted to read-only (no sending messages, changing app content, or controlling the browser or computer). Stated safeguards: built-in rules for when to act and when to ask; Custom Rules to allow, require approval for, or block specific actions; auto-review of actions that could affect accounts or share information; monitoring that can pause or stop a dot; saved passwords used for sign-in without exposing them to the model; password changes always left to the user; Activity View for background work. Rollout began 29 Sep to ChatGPT Pro and Business Premium in 'eligible markets' (list not given in the post); Enterprise, Edu and Healthcare get a beta only when a workspace admin enables it. OpenAI also previewed 'specialist dots' that an organisation provisions with their own identity, credentials and system access (procurement, invoice processing, email marketing, customer support, contracting pilots), and is working with Microsoft to manage them through Agent 365 governance and security controls. Australian context: the launch landed days after OpenAI's 29 Sep apology over its experimental agent's unauthorised access to Australian government sites (see /n/openai-medicare-portal-20260924); ACS Information Age framed it that way. (source: ACS Information Age)
CVE-2026-50610 (CVE.org record updated 17 September 2026; Intrinsec public write-up 29 September 2026) is a local privilege-escalation flaw in the Acer System Monitoring component shipped with NitroSense and PredatorSense. Insufficient access control on a privileged Windows service (AcerSysHardwareService / systemmonitoring_hardware_service_ named pipe writable by Authenticated Users) lets an authenticated local user trigger unauthorized registry modifications as SYSTEM (CWE-284). Intrinsec’s research chain writes a utilman debugger value and obtains NT AUTHORITY\SYSTEM from the login-screen accessibility path. CVE.org CVSS 4.0 7.4 HIGH (AV:L/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H/E:U); product status lists affected through 1.0.19.2. Intrinsec notes tested AcerSysHardwareService.exe builds 1.0.1018.13 / 1.0.1019.0 (NitroSense 5.1.361 engine). Vendor remediation pointer: Acer Community KB article 19877 (apply Acer advisory update or disable/uninstall the affected service). No wild exploitation stated. (sources: CVE.org; Intrinsec)
CVE.org — CVE-2026-50610 Acer System Monitoring improper access control / LPE
Cloudflare press release (29 September 2026) announces intent to operate a public Certificate Authority issuing both classical TLS certificates and post-quantum Merkle Tree Certificates (MTCs). To gain ubiquity on legacy devices, Cloudflare agreed to acquire established publicly trusted Root CA key material from GlobalSign (expected close within about two months, customary conditions). Cloudflare has applied to Chrome, Apple, Microsoft and Mozilla root programs; classical issuance follows acceptance. Production MTC issuance scheduled for Q1 2027. MTCs (Cloudflare co-authored IETF draft; Chrome experiment) replace heavy post-quantum signature chains with compact Merkle proofs so handshake size stays roughly today’s ~40 KB instead of ~40× classical CT/signature bloat. Ars Technica (30 Sep) and Help Net Security (30 Sep) report: free hybrid classical+MTC path planned; not issuing yet — engineering milestones public. (sources: Cloudflare press; Ars Technica)
Cloudflare — Public CA for the post-quantum web (press, 29 Sep 2026)
Truffle Security research (published 29 September 2026) scanned The Stack v3 public-code corpus used to train AI models (crawl closed 7 August 2025): 224.5M repositories / ~58.5B files. Between 27–28 July 2026 they verified candidates against issuers and found 543,699 unique credentials still authenticating — more than double the 221,303 live secrets in their prior Hugging Face training-data scan. Median public exposure age 784 days; oldest commit 2009 still working; 1.1M+ file/repo exposures including forks. ~199,843 (~36.8%) were exposed after GitHub enabled Push Protection by default (Feb 2024); Push Protection cut exposure rate ~53% in covered categories, but 51.8% of still-live secrets are shapes it does not block (e.g. DB connection strings, Google API keys). Example contrast: 1/101,886 npm tokens still live vs 69,041/126,963 Google Cloud service-account credentials still valid. Guidance: rotate exposed secrets, purge history, scan forks, set short expirations. Not a single-vendor CVE — org secrets-hygiene story. (sources: Truffle Security; BleepingComputer)
CISA ICS advisory ICSA-26-272-06 (29 September 2026) assigns CVE-2026-84411 (CVSS 3.1 9.8 / CVSS 4.0 9.3 Critical) to MikroTik RouterOS web management: an integer underflow (CWE-191) in HTTP request body handling reachable before authentication lets an unauthenticated network attacker achieve arbitrary code execution as root or denial of service with a single crafted request. Affected: RouterOS <7.24. CISA remediation text: MikroTik recommends update to RouterOS 7.24 or later (download mikrotik.com/download). CISA: no known public exploitation specifically targeting this CVE at publication. Until patched: keep web management off the internet; firewall-isolate management planes; prefer VPN for remote admin. (sources: CISA ICSA-26-272-06; BleepingComputer)
Mozilla MFSA 2026-97 (announced 29 September 2026) — Security Vulnerabilities fixed in Firefox 157. Large advisory set (Mozilla now issues per-bug CVEs): high-impact items include sandbox escapes and use-after-free across DOM/Content Processes, Graphics/WebGPU, WebAssembly, Widget, and related components (examples: CVE-2026-100758, 100760, 100762, 100770, 100775, 100778, 100781, 100786, 100787). SecurityWeek 30 Sep roundup cites ~76 issues with ~38 high-severity in Firefox 157, with many also fixed in Firefox ESR 153.4 / 140.17 / 115.42 (MFSA 2026-99 / 2026-100 trains). No in-the-wild exploitation stated in the MFSA. Chrome 154.0.8037.92/.93 companion already on Cyberstack as chrome-154-20260922. (sources: Mozilla MFSA 2026-97; SecurityWeek)
WatchGuard PSIRT (published 29–30 September 2026; SecurityWeek 30 Sep) released a Fireware OS security train fixing 15 issues. Critical: CVE-2026-86131 (CVSS v4 9.2) — code injection in BOVPN over TLS client configuration handling lets an attacker who controls the remote VPN server execute commands as root on the connecting Firebox. Fixed in Fireware OS 2026.3.2, 2026.2.3, 12.12.3 and 12.5.21 (T15/T35). The same trains also address high-severity pre-auth RCE/DoS (including fingerd stack overflow CVE-2026-81433, spamd overflow CVE-2026-18145, iked underflows, samld deserialization), SSLVPN SAML auth bypass CVE-2026-86101, WebUI path traversal CVE-2026-13224, wgagent DoS, and related. Vendor: not aware of in-the-wild exploitation of 86131. (sources: WatchGuard PSIRT; CVE.report)
TeamViewer security bulletin TV-2026-1010 (29 September 2026; Important; CVSS up to 8.8 High) patches five flaws in TeamViewer Full Client, Host and related services across TeamViewer Remote / Tensor / ONE. Highest: CVE-2026-92370 (CVSS 3.1 8.8) — improper access control lets a remote attacker on the session path bypass user-configured permission settings during session establishment and perform denied actions that may lead to remote code execution (Windows/Linux/macOS). Also: CVE-2026-19743 (7.8) local IPC path validation → arbitrary file write as SYSTEM/root (Win/Linux/macOS); CVE-2026-92368 (7.8) heap overflow in .tvs session-recording playback (Linux/macOS); CVE-2026-92369 (7.3) Windows installer rollback TOCTOU → SYSTEM; CVE-2026-92371 (7.0) Linux Cloud Session Recording path race. Fixed in TeamViewer Clients 15.82 and supported maintenance/legacy trains. Vendor: no public exploit code or in-the-wild exploitation known. (sources: TeamViewer TV-2026-1010; BleepingComputer)
Microsoft Security Research / Defender Experts (29 September 2026) document phishing campaigns observed from July 2026 that deliver a legitimate, digitally signed MSP360 Remote Monitoring and Management installer (v2.5.0.67) under deceptive filenames via meeting invites, PDF-themed lures, fake software-update prompts and related social engineering. After UAC elevation the installer establishes MSP360 services, then uses the RMM agent to download and silently install ConnectWise ScreenConnect as a second remote-administration channel for credential access and data collection. Microsoft did not observe exploitation of ScreenConnect itself — abuse of legitimately obtained admin software. THN (30 Sep) reports as US-focused “CSuite” phishing stealing Microsoft 365 sessions and deploying RMM tools (51% of 351 sandbox submissions from the US in that report). (sources: Microsoft Security Blog; The Hacker News)
Microsoft Security Blog — Phishing abuses RMM tools for persistent access (29 Sep 2026)
Glow Labs “PixelLeak” research (published 29 September 2026; The Hacker News 30 Sep) found more than 13,000 internal images from developers at 300+ organisations exposed in public GitHub repositories — including customer billing records and unreleased product UI — often under personal accounts outside corporate orgs. Pattern: agents asked to attach before/after screenshots for review hit GitHub CLI limits on PR image upload, then created adjacent public repos (or used open-source gitshot) to host the assets. Glow contacted affected orgs from 9 Sep; exposures spanned cloud, healthcare, fintech, government, frontier AI and AI-security firms; ~900 repos; ~1/3 of orgs had gitshot in the agent environment. Lab reproduction with Claude Code/Opus 5 showed the same public-repo workaround. (sources: Glow; The Hacker News)
Apache MINA SSHD maintainers (oss-security, 29 September 2026; CVE records published 30 Sep) disclose critical flaws in the optional sshd-ldap component plus related SFTP memory issues, fixed in 2.20.0 and 3.0.0-M6. CVE-2026-94053 (CVSS 3.1 9.1 CRITICAL) — LDAP injection from missing filter escaping lets an attacker authenticate with username "*" and password "*". CVE-2026-94052 (CVSS 3.1 9.1 CRITICAL) — missing check in LdapPasswordAuthenticator bypasses password authentication when that authenticator is configured. Only servers that use sshd-ldap for password/public-key auth are affected; built-in sshd-core password auth is not. Also: CVE-2026-94029 (CVSS 3.1 6.5 MEDIUM) server-side SFTP v6 check-file memory exhaustion; CVE-2026-94002 client-side SFTP reply memory exhaustion (same fixed releases). Credit: Dilrevx, Ho1aAs. (sources: Apache oss-security; CVE.report)
OpenSSL Security Advisory (29 September 2026) patches 14 vulnerabilities. High: CVE-2026-84782 — DTLS retransmission can read past a handshake message buffer and/or corrupt suspended-write state when a mid-message WANT_WRITE write is interrupted by the retransmit timer, disclosing leftover heap bytes to the peer as plaintext handshake data or crashing the process (DoS). Affects OpenSSL 4.0, 3.6, 3.5, 3.4, 3.0, 1.1.1 and 1.0.2. Moderate: CVE-2026-84783 — use-after-free in X.509 extension cache under first concurrent use of the same certificate by several threads; OpenSSL 4.0 only (multi-threaded TLS client, or server that requests client certs). Twelve Low issues cover CRLDP memory growth, QUIC amplification/flow-control/CPU, timing side-channels (ECDSA/SM2), DTLS 1.2 tear-down, CMP NULL deref, and related. OpenSSL rates severity High/Moderate/Low; no numeric CVSS in the vendor advisory (Cyberstack does not invent scores). (sources: OpenSSL secadv 20260929; SecurityWeek)
