60
Smart search ranks by meaning as well as keywords (one row per story, last 45 days).
Researchers Uncover RovoBlast Vulnerability in Atlassian AI Assistant
Atlassian fixed RovoBlast, a flaw letting a single crafted link make its Rovo AI assistant exfiltrate company data.
Researchers disclosed a vulnerability dubbed RovoBlast affecting Atlassian's Rovo AI assistant. A single crafted link could cause the assistant to exfiltrate company data accessible to it. Atlassian has since fixed the flaw; no widespread exploitation was reported in the disclosure.
55
Microsoft Copilot Personal Flaws Could Let One Click Exfiltrate Data From Connected Apps
Varonis discloses CoSnitch (CVE-2026-24301), three Microsoft Copilot Personal flaws enabling one-click exfiltration of connected-app data; patched August 18, 2026.
Varonis Threat Labs found that an undocumented autorun=1 parameter, paired with the q parameter, lets an attacker-supplied prompt run automatically on page load in a victim's authenticated Copilot session, then exfiltrate data from connected services such as mail, calendar, Google Drive, chat history and the memory store via Copilot's built-in URL fetch to an attacker webhook. A separate memory-poisoning path through web summarization lets a crafted page persist attacker instructions in the user's memory, surviving password changes, session revocation and device re-enrollment. Microsoft shipped patches on August 18, 2026, tracked as CVE-2026-24301, and Varonis found no evidence of in-the-wild exploitation. The flaws were found via 'meta-hacking', asking Copilot itself to reveal the autorun parameter and its protections.
66
ThinkPrior: Zero-Rollout Difficulty Priors for Cold-Start Prompt Selection in RLVR
ThinkPrior builds zero-rollout difficulty priors via an offline verifier-anchored pass, halving silent groups in RLVR and cutting wasted rollouts on Qwen2.5-Math-7B.
In GRPO-based RLVR, groups where all rollouts are correct or all are wrong yield zero advantages and consume about 39% of a run's rollouts under uniform sampling. ThinkPrior initializes a Beta posterior from an external anchor pass's verifier-scored pass rate, selecting prompts by expected learnability before any target-policy rollout, without changing the loss or optimizer. On Qwen2.5-Math-7B across sixteen seeds it more than halves early silent groups and cuts wasted rollouts through step 30 by nearly a fifth, with no detected final-accuracy difference. The ThinkPrior+DAPO composition reduces generated rollouts by 10.6% at an equal 3,840-rollout update budget.
29
Gradium Launches Voice Design: Write a Prompt, Get a Brand New Synthetic Voice in Seconds
Gradium, a Kyutai spinout, launched Voice Design, generating custom synthetic voices from text descriptions in seconds across five languages.
Gradium, a Paris-based voice AI company spun out of Kyutai, launched Voice Design, which generates new synthetic voices from 1-500 character text descriptions in seconds without needing reference audio or speaker consent. The feature is live in the Gradium API and Studio, free on every plan including the free tier, and kept voices run on the standard streaming TTS endpoint at the same latency as catalog voices. Vendor-run blind pairwise listening tests across 7,627 comparisons report a 72.6% win rate, 13.6 points ahead of ElevenLabs at 59.0%, placing first in all five tested languages, with the largest margins on regional accents such as Quebecois French (97%).
32
DataFlex-RL: An Evaluation Platform for RLVR Data Policies
DataFlex-RL benchmark of 13 RLVR data policies on Qwen2.5-7B finds none reproducibly beats uniform sampling under matched GRPO training.
DataFlex-RL is an evaluation platform comparing rollout-selection, reweighting, and mixture data policies for RLVR under a common GRPO recipe. Across 13 configurations and 12 matched seeds with Qwen2.5-7B-Base on 12 math, logic, and science benchmarks, uniform GRPO improved domain-balanced accuracy by 7.76 points, but no alternative policy achieved a statistically significant improvement. A corrected 12-seed Llama-3.1-8B-Base extension found no consistent winner, and math-heavy evaluation summaries were negatively correlated (-0.33) with domain-balanced summaries.
35
Building a Production Greek-English Speech Recognizer
Engineering report details Sophea, a production Greek-English ASR reaching 4.26% WER on public English sets via ROVER ensemble and data-pipeline calibration.
Across 23 training iterations, two architectures, and nine production gates, no single data composition passed all gates; a three-model ROVER ensemble reached 9 of 9 gates and cut overlapping-speech WER from 53.35% to 37.87%. Calibrating an audio-quality filter against in-domain anchors reduced discarded scored Greek audio from 98.7% to 10.6%, and a pre-registered ablation traced a hallucination defect to one training-data package. The sophea/asr-k1 preview arbiter lists 4.26% average WER on eight public English test sets and 25.88% WER on live Greek noisy traffic; no weights or training data are released.
32
Eliciting Weak-to-Strong Generalization with On-Policy Reverse Distillation
OPRD distillation enables weak-to-strong generalization by amplifying verifier-supported policy updates, outperforming existing RL and distillation methods with fewer student updates.
On-Policy Reverse Distillation (OPRD) evaluates a weak teacher's policy shift relative to its reference policy on student rollouts and amplifies the verifier-supported component of the student's policy gradient. This rescaling preserves the stationary points of policy optimization while letting the student learn beyond the teacher's capacity ceiling. In successive model transfer and multi-teacher distillation, OPRD achieves higher performance with fewer student updates than existing RL and distillation approaches, and response-style analysis shows students remain closer to verifier-RL-trained models than to their weak teachers.
33
Attackers Exploit Critical Switchvox Flaw to Deploy Reverse Shells Without Credentials
Attackers exploit unauthenticated SQL injection CVE-2026-9586 in Sangoma Switchvox to run PostgreSQL commands and deploy reverse shells.
Threat actors are exploiting CVE-2026-9586 (CVSS 9.3), an unauthenticated SQL injection in Sangoma Switchvox SMB Edition 8.3 (104997), since August 30, 2026, running arbitrary SQL as the PostgreSQL superuser and achieving remote code execution. The /pa endpoint concatenates the user-controlled PhoneIP value into PostgreSQL queries; attackers can extract database contents, escalate to Switchvox web administrator, exfiltrate the cookie signing key to forge authentication, and invoke reverse shells. Sangoma patched the flaw in Switchvox 8.4.0.2 on July 14, 2026, roughly 4,000 instances are internet-exposed (mostly in the US), and honeypot activity from IP 176.65.148.184 deploys reverse shells followed by Base64-encoded process enumeration.
70
Fine-tuning a 350M Model for Better Structured Outputs in 100 GRPO Steps
Hugging Face guide fine-tunes a 350M-parameter model with 100 GRPO steps to improve structured output reliability.
A Hugging Face blog post demonstrates fine-tuning a 350M-parameter model using GRPO (Group Relative Policy Optimization) with TRL over 100 training steps. The stated goal is more reliable structured outputs from small language models. No article body was available, so details beyond the title are limited.
22
US, UK, Dutch Agencies Expose Iranian ‘Chosen Brick’ Surveillance Malware
US, UK, and Dutch agencies warn Iranian state actors deploy Windows surveillance malware 'Chosen Brick' against dissidents, activists, and journalists worldwide.
Joint advisories from US, UK, and Dutch agencies describe Chosen Brick, a Windows surveillance malware active since at least 2025 and used by Iranian state cyber actors to track regime opponents. The malware harvests contacts, emails, and social media messages, persists via registry Run keys, evades Microsoft Defender, and uses per-victim Telegram bot IDs for command-and-control and exfiltration. Operators build rapport on WhatsApp and Telegram posing as acquaintances or support staff, disguising payloads as utility software or fake medical documents. Capabilities include screenshot capture, audio recording, credential theft, secondary payload delivery, and data wiping.
60
A Princeton Researcher Proposes Recurrent Looped Transformer (RLT) that Carries Decoder State across Every Token, Fixing 96 Blocks per Token with Unbounded Temporal Depth
Princeton researcher Yifan Zhang proposes Recurrent Looped Transformer, carrying full decoder state across every token for unbounded temporal depth.
Yifan Zhang's technical report defines the Recurrent Looped Transformer (RLT), pairing a causal encoder with a recurrent decoder whose final output and layerwise sliding-window attention cache carry into every subsequent token with no prompt-response boundary reset. The reference configuration ties 48 encoder and 48 decoder layers, executing 96 logical blocks per token while the state path grows to 48t blocks after t tokens at fixed per-token compute. The report details RL replay contracts that rebuild all states under current parameters and exact prefix snapshots for multi-turn serving, but explicitly reports no measured efficiency, reasoning quality, or scaling results.
40
Iranian cyber targeting of dissidents, activists and journalists
UK NCSC, FBI, and Dutch AIVD expose CHOSEN BRICK spyware used by Iranian state actors against dissidents, activists, and journalists worldwide.
A joint advisory from the UK NCSC, FBI, and Dutch AIVD details CHOSEN BRICK, a Windows spyware family used by Iranian state cyber actors since at least 2025 against dissidents, activists, and journalists in the UK, US, and Netherlands. Actors build rapport on WhatsApp and Telegram impersonating known contacts or platform support, then deliver disguised payloads resembling apps such as Telegram, Norton, RunwayML, or fake MRI results. The malware persists via HKCU Run registry keys, adds Microsoft Defender exclusions, and uses a unique Telegram bot C2 per victim. Capabilities include screen capture, microphone recording, process enumeration, email and messaging data theft, file deletion, and system wiping; victim data has appeared on pro-Iranian leak sites.
78
Seeing is Not Believing: Breaking the Physical-to-Digital Trust Boundary in Robotics
Researchers show a single ROS 2 environment variable lets attackers inject fake telemetry and hijack robots while spoofing downstream remote attestation.
A pre-built hook loaded via one modified environment variable covertly intercepts and injects both telemetry and control signals before publication in ROS 2, breaking the physical-to-digital trust boundary in multi-robot task handovers. Attackers can also distribute compromised third-party Docker containers and auxiliary tools embedding the hooks. On a physical Franka Emika arm running Secure ROS 2, the attack injects fabricated telemetry in real time with roughly 3 ms jitter and achieved an 87% success rate even against an AI-based detector. Findings were responsibly disclosed to the ROS 2 development team.
50
Risky Bulletin: Slovakia finds Russian backdoor in traffic speed cameras
Slovakia's NBU found an SMS-triggered backdoor in Russian-made NERO R-ONE traffic cameras, pausing a 279-unit deployment.
Slovakia's national security service NBU issued an alert against NERO R-ONE high-speed traffic cameras after finding a backdoor that grants shell and network access via SMS from hardcoded Russian phone numbers. The cameras are a rebranded version of the Russian CORDON PRO.M model by St. Petersburg firm Semicon, purchased via a Cyprus shell company under a €30 million EU-funded project. The report also found SecureBoot disabled, vulnerable web management, and unauthenticated live streams; the Interior Ministry paused deployment of 279 cameras pending independent assessment.
72
Off the Hook: Discovering and Observing Active Exploitation of Sangoma Switchvox CVE-2026-9586
Horizon3 disclosed CVE-2026-9586, an unauthenticated SQL injection in Sangoma Switchvox leading to RCE, now under active exploitation in the wild.
Horizon3.ai attack researchers discovered CVE-2026-9586, an unauthenticated SQL injection vulnerability in Sangoma's Switchvox VoIP appliance that can escalate to remote code execution. The researchers observed active exploitation of the flaw in the wild. A disclosure write-up was published alongside their findings, and defenders should treat internet-exposed Switchvox instances as at risk.
72
Zero-Click Grok Chat History Theft: Adversa AI Demonstrates Cryptographic Context Injection
Adversa AI's Cryptographic Context Injection bypasses AI guardrails using AES-encrypted payloads, enabling zero-click theft of Grok users' full chat histories.
Adversa AI researcher Rony Utevsky disclosed Cryptographic Context Injection, which hides instructions in AES-256-GCM ciphertext and tricks models into decrypting them inside their own code execution runtime, where the output is treated as trusted. Demonstrated against xAI's Grok, it stole user names, locations, subscription tiers, and full chat histories with zero clicks, and against Google's Gemini to bypass safety rules, generate incendiary-device instructions, and expose system instructions. Reported to xAI on June 3, 2026, the Grok attack remained reproducible as of August 19, 2026; the Gemini issue was not formally reported because Google's bug bounty excludes jailbreaks.
68
DeepZero: Open-source hunting for vulnerable Windows drivers
DeepZero, a new open-source engine, automates discovery of exploitable Windows kernel drivers for BYOVD attacks using Ghidra, Semgrep, and an LLM.
DeepZero is a free, open-source Python pipeline orchestrator that automates hunting for exploitable Windows kernel drivers relevant to BYOVD (bring your own vulnerable driver) attacks. Its seven-stage YAML pipeline parses PE headers, filters for kernel-mode drivers with IOCTL surfaces, excludes drivers listed on loldrivers.io, then runs headless Ghidra decompilation, Semgrep scanning, and an LLM-based exploitability assessment. The maintainer reports multiple verified vulnerabilities in the Snappy Driver Installer corpus, some still in the disclosure process, and notes findings involving plug-and-play-created device objects may need physical hardware to confirm.
45
Grok exfiltrates user data when malicious instructions are encrypted
Researchers show Grok can be made to exfiltrate user data via Cryptographic Context Injection, a newly documented technique that bypasses LLM safety guardrails.
According to Ars Technica, Grok exfiltrates user data when malicious instructions are encrypted, a technique called Cryptographic Context Injection. The method is described as the latest documented way to break LLM safety guardrails, showing that encrypted content can carry hidden instructions past safeguards. The finding underscores gaps in how large language models validate and execute context from external sources.
55
DriveZero: End-to-End Driving Beyond Human Demonstrations
DriveZero pairs a frozen vision-foundation-model perception stack with a PPO-trained closed-loop RL teacher to beat replay experts on nuPlan.
DriveZero is an end-to-end camera-only autonomous-driving planner that separates perception and action. Its DriveVFM perception backbone consolidates frozen vision foundation models (DINOv3, SigLIP2, SAM, Depth Anything V2) from raw images without task annotations, while DriveRL trains a privileged PPO teacher policy through closed-loop rollouts in interactive worlds built from real driving logs. The planner distills this teacher, achieving a 93.57 mean nuPlan score across Val14, Test14-hard and Test14-random splits and beating the Log-Replay expert on all three. It also sets state of the art on NAVSIMv1, NAVSIMv2 and closed-loop HUGSIM without human trajectory supervision.
30
Encoded Early, Used Late: Where Transformers Begin to Act on an Inferred Partner's Expertise
Probing finds transformers represent an inferred dialogue partner's expertise in early layers long before it causally influences output.
Using ExpertCollab, a corpus of multi-turn research-planning dialogues between model-played personas at four expertise levels, researchers show that a partner's inferred expertise is most decodable in early transformer layers and decays to near chance before the network's midpoint. Counterfactual patching reveals that injecting the expertise difference at peak decodability barely changes a fixed late-layer readout, while injection past the midpoint propagates almost completely. The result bounds where readout or steering of partner-conditioned behavior must intervene, demonstrated on a single model with a synthetic corpus.
30
New AI Attack Hides Malicious Instructions in Normal-Looking Text to Evade Safety Filters
Check Point researchers show crafted prose hides policy-violating instructions that bypass all tested LLM gatekeepers, including GPT-4o mini and Llama Guard 3.
A new prompt-crafting technique embeds malicious payloads inside grammatical, natural-looking text without Base64, invisible Unicode, or obvious encodings, defeating lightweight pre-screening gatekeepers. In testing, all four evaluated gatekeeper models—gpt-4o-mini-2024-07-18, gpt-oss-safeguard:20b, claude-3-haiku-20240307, and llama-guard3:8b—classified the crafted wrappers as safe at a 100% bypass rate across 23 obfuscated prompts. GPT-5 Thinking in high-reasoning mode recovered and acted on the hidden instruction in 17 of 18 tests (~94.4%), often spending over a minute and multiple Python executions. Researchers recommend paraphrasing untrusted input, hardening gatekeeper policies, and applying defense-in-depth controls for agentic deployments.
50
HOL Guard: Open-source antivirus for AI agents
HOL Guard is an open-source local guardrail that pauses AI coding agents before risky actions like secret access and prompt injection.
HOL Guard sits between AI coding agents (Claude Code, Cursor, Codex, Gemini CLI and others) and the host machine, intercepting risky commands before execution with checks taking under 50 milliseconds and running fully offline. It offers four sensitivity modes — Gentle, Balanced (default), Strict, and Paranoid — and parses command structure, environment, sensitive-path access and network destinations to decide when to interrupt. The core runtime is free and open source on GitHub, with 552,000 downloads reported; the vendor says it has no telemetry on adoption because collection is off by default.
32
What We Missed: Did ShinyHunters 'Breach' ReliaQuest?
Dark Reading editors discuss whether ShinyHunters breached ReliaQuest and new research questioning the prevalence of AI-generated malware.
Dark Reading editors review stories they had not previously covered in a video discussion, centered on recent activity attributed to the ShinyHunters threat actor and whether it constitutes a breach of security services firm ReliaQuest. The conversation also touches on new research about how common AI-generated malware actually is. No new indicators, victims, or technical details are provided beyond the discussion format.
25
Datamimic – don't let your coding agent invent its own test world
Datamimic is an open-source test data generation tool aimed at keeping coding agents from inventing their own test fixtures.
A Hacker News discussion (40 points) highlights Datamimic, an open-source rapiddweller GitHub project for generating realistic synthetic test data. The tool targets AI coding agents, aiming to prevent them from fabricating their own inconsistent test worlds. Only the repository link was shared, so details are limited.
15
Iranian spies hit Windows machines with Chosen Brick data-stealing malware
FBI, UK NCSC, and Dutch AIVD warn Iranian intelligence uses Chosen Brick spyware against dissidents, stealing contacts, emails, and messaging data.
A joint advisory from the FBI, UK NCSC, and Dutch AIVD says Iranian state cyber actors have used the Chosen Brick Windows malware since at least 2025 to surveil dissidents, activists, and journalists. Attacks begin with heavily researched WhatsApp and Telegram messages impersonating trusted contacts, tricking victims into opening fake installers resembling Pictory, RunwayML, Norton Antivirus, Telegram, Adobe Flash Player, and KeePass. The malware persists via the HKCU Run registry key, adds Microsoft Defender exclusions, uses victim-specific Telegram bots for C2, captures screen and audio, steals emails and Telegram/WhatsApp data, and can wipe systems.
68
DPRK APTs: Ted backdoor and curlRAT target South Korean media and automotive sectors
Rapid7 uncovered a DPRK-linked Linux toolkit using a HAProxy-embedded ted backdoor, SSH keylogger, and curlRAT against South Korean media and automotive firms.
Rapid7 Labs identified a previously undocumented framework attributed with medium confidence to DPRK actors, targeting South Korean automotive and media organizations likely since early 2025. The toolkit embeds a backdoor compiled into HAProxy 2.8.12 using its filter API, plus trojanized crond, agetty, atd, sshd, and polkitd, an SSH keylogger storing credentials under /var/lib/sshd/, and a curl-based RAT with a watchdog thread. It enables remote command execution, malicious script injection into served webpages (a watering-hole loop), credential harvesting, and long-term surveillance. Hardcoded C2s are associated with APT37 via ThreatFox, and exposed groupware portals and mail servers align with Kimsuky tradecraft; the initial access vector and any CVE remain unconfirmed.
68
North Korea-linked Hackers Hide a Backdoor Inside HAProxy
Rapid7 reports North Korea-linked hackers implanted a backdoor compiled into HAProxy at South Korean automotive and media firms, enabling covert C2 and credential theft.
Rapid7 documented a previously undocumented Linux toolkit hitting South Korean automotive and media organizations, centered on a backdoor compiled directly into victims' HAProxy 2.8.12. The 'ted backdoor' uses HAProxy's native filter API to intercept HTTP traffic, receive C2 commands hidden in requests to a fake image path, and erase all traces from logs and counters; the toolkit also trojanizes crond, agetty, atd, sshd, and polkitd, adds an SSH keylogger, and runs curlRAT with virtualization checks. It can inject scripts or replace page content for selected victims, turning the load balancer into a watering hole. Attribution sits at medium confidence toward North Korean state actors, with overlaps to APT37-linked infrastructure and a concurrent Lazarus campaign; the campaign's command domains have since gone dark.
60
What Matters, When? Diagnosing and Improving Conditional Visual Grounding in Visuomotor Imitation Policies
Researchers diagnose conditional visual grounding failures in visuomotor imitation policies and show targeted interventions substantially improve distractor robustness.
The paper studies why ACT-based visuomotor imitation policies fail when visually similar distractor objects or receptacles are introduced, finding sensitivity depends on both distractor type and manipulation stage. Interventions including distractor augmentation, phase-dependent attention regularization, and appearance-based visual prompting improve target selection while preserving spatial control information, with gains in simulation and on a physical UR3e. The same failure pattern is confirmed in a pretrained vision-language-action policy on a state-conditioned medical instrument-handling task.
18
ECHO: A Matched-Contrast Benchmark for Context-Sensitive Turn-Taking in Full-Duplex Dialogue
ECHO benchmark pairs identical-overlap Chinese dialogue examples with contrasting contexts, revealing most full-duplex systems bias toward yielding the floor.
ECHO is a paired diagnostic benchmark for Chinese full-duplex turn-taking that matches examples with identical overlap transcripts but contrasting preceding multi-turn contexts, one requiring Yield and the other Keep. It also includes off-talk examples for diagnosing unnecessary yielding and introduces pair accuracy, which grants no credit to constant-action policies. Experiments across multiple full-duplex systems show most exhibit a pronounced Yield bias, performing substantially better on interruptions than backchannels. The benchmark and metadata will be publicly released.
40
NeoMME: an efficient Multimodal-native and Multilingual Encoder
H Company released NeoMME, an efficient multimodal-native and multilingual encoder, via a post on the Hugging Face blog.
H Company published a Hugging Face blog post introducing NeoMME, described as an efficient multimodal-native and multilingual encoder. The article body was unavailable in the feed, so architecture, benchmarks, and licensing details could not be verified. The post suggests a new encoder release relevant to multilingual multimodal model development.
32