Search: “usage-limits”

Smart search ranks by meaning as well as keywords (one row per story, last 45 days).

ZDI-26-668: Adobe Acrobat Reader DC Annotation Use-After-Free Information Disclosure Vulnerability

ZDI advisory ZDI-26-668 reports an annotation use-after-free (CVE-2026-81984) causing information disclosure in Adobe Acrobat Reader DC.

The Zero Day Initiative published advisory ZDI-26-668 for a use-after-free condition in the annotation feature of Adobe Acrobat Reader DC. Exploitation allows remote attackers to disclose sensitive information when the target opens a malicious file or page. ZDI rated the issue 3.3 on the CVSS scale and assigned CVE-2026-81984.

ZDI Published Advisories·Sep 10, 05:00 UTC · 6d agoAdvisory 19 sourcesCVE-2026-81984

15

Anthropic Adds Plugin Evals to Claude Code: 6 Grader Types, a No-Plugin Baseline, and a CI Gate for Skills

Anthropic ships a plugin evals workflow for Claude Code with six grader types, a no-plugin baseline arm, and a CI gate via threshold and cost flags.

Anthropic published a plugin evals workflow for Claude Code, exposed via the "claude plugin eval" command on v2.1.269+. Six grader types exist: regex, tool_used, tool_order, and file_exists are free transcript checks, while llm and baseline invoke a billed judge model. Every case runs with and without the plugin, and the delta (Δ) isolates the plugin's contribution; a Δ near zero with a failing tool_used:Skill grader indicates the skill never triggers. CI gating uses --threshold 0.8, --max-cost-usd, --trust-plugin, and --no-publish flags, with results written to a report.html under evals/results/.

MarkTechPost·Sep 11, 21:05 UTC · 4d agoAI tools & infra2

42

USN-8736-2: Perl vulnerabilitiesnew

Ubuntu issued USN-8736-2 fixing two Perl regex flaws that could cause out-of-bounds heap access, denial of service, or security bypass on 24.04 LTS.

USN-8736-2 backports the Perl fixes from USN-8736-1 to Ubuntu 24.04 LTS. CVE-2026-15534 involves out-of-bounds heap reads or writes when regular expressions handle large inputs, potentially causing denial of service or arbitrary code execution. CVE-2026-19487 causes incorrect regex matching with alternative branches, allowing security restrictions to be bypassed.

Ubuntu Security Notices·14:37 UTC · 4h agoAdvisory 13 sourcesCVE-2026-15534CVE-2026-19487

28

Structural Jailbreaks Generalize but Do Not Compound: A cross-provider and multilingual study of Involuntary In-Context Learning

Researchers show IICL structural jailbreaks generalize to Google Gemini, lifting attack success to 80-100% on harm and financial benchmarks; non-English prompts attenuate it.

The study red-teams two Google Gemini models with Involuntary In-Context Learning (IICL), a structural jailbreak reframing harmful requests as the final cell of a data-labeling task. IICL lifts attack success from at most 6.7% to 80-90% on HarmBench and 97-100% on financial abuse (FinProof), an order of magnitude above prior results on OpenAI's GPT-5.4. Against a compounding hypothesis, forcing IICL output into Spanish, Hindi, or Arabic attenuates the attack in 11 of 12 conditions, attributed to a 'relevance curse' producing lower-quality harmful content in lower-resource languages. Findings replicate under an independent non-Google judge (Cohen's kappa 0.86 over 377 paired verdicts).

arXiv cs.CR·Sep 8, 07:38 UTC · 8d agoAI safety & security

58

Safety for Whom? Boundary-Aware Self-Distillation for Controlled LLM Safety Refusal

A self-distillation safety framework tunes narrow-boundary refusals in Qwen3-8B, raising target-domain refusal to 84.75% while cutting over-refusal from 15.20% to 5.20%.

The paper formulates narrow-boundary safety, where deployments need refusals within specific topics rather than whole subjects, and proposes an offline self-generated framework with controlled topic generation, escalating retries, and harmful-benign boundary pairs. On political persuasion with Qwen3-8B, the method raised target-domain refusal from 9.47% to 84.75% and cut the mean unsafe-response rate across three broader benchmarks from 26.26% to 0.14%. Verified target-model responses reduced over-refusal from 15.20% to 5.20%, and boundary-pair data cut comply-side over-refusal on held-out pairs from 32.94% to 4.16%. Results show data composition controls the safety-usability trade-off and alignment should be evaluated on both sides of the refusal boundary.

Hugging Face daily papers·Sep 2, 20:00 UTC · 13d agoAI safety & security1

32

Bad Likert Judge: A Novel Multi-Turn Technique to Jailbreak LLMs by Misusing Their Evaluation Capability

Unit 42 details the Bad Likert Judge multi-turn jailbreak that abuses LLMs' evaluation capability, raising attack success rates over 60% across six frontier models.

Palo Alto Networks Unit 42 describes the Bad Likert Judge technique, a multi-turn jailbreak that asks a target LLM to act as a Likert-scale judge scoring the harmfulness of example responses. The highest-rated example in each scale can carry harmful content, bypassing the model's internal guardrails. Testing across six state-of-the-art text-generation LLMs showed an average attack success rate increase of more than 60% versus plain attack prompts, with tested models anonymized. The technique targets edge cases rather than typical use, and the article positions the work as guidance for defenders on potential jailbreak risks.

Palo Alto Unit 42·Aug 17, 12:34 UTC · Aug 17, 2026AI safety & security

60

Trends in Web Threats: Attackers Were More Active During Holiday Season

Unit 42 tracked 533,000 malicious landing URL incidents from October-December 2021, showing web threats peaked during the holiday shopping season.

Unit 42 detected 533,452 malicious landing URL incidents (120,753 unique) and 2,906,875 malicious host URL incidents (165,255 unique) from October through December 2021. Threat activity peaked in November, likely tied to Black Friday in the United States, United Kingdom, and Germany. Most malicious domains appeared to originate in the United States, followed by Russia and Germany. Personal sites, blogs, business sites, and shopping sites were the most common apparently benign entry points for attacks.

Palo Alto Unit 42·Aug 17, 13:06 UTC · Aug 17, 2026Research

15

Can We Stop The Ads? Taxonomy and Characterization of Smartphone Splash Ads and Existing Countermeasures

Study of 108 ad-defense implementations finds only one tool blocked splash-ad navigation across ten popular apps, and it required Accessibility permission.

The paper taxonomizes smartphone splash ads — full-screen ads at app launch that trick users into trigger mechanisms such as moving the phone — and analyzes 108 documented advertising defenses for deployment barriers. Many defenses require device rooting, jailbreaking, runtime code injection, or application modification; others need extra permissions, rule maintenance, compilation, or payment. In evaluating 13 configurations of 11 tools across 10 popular apps, only one prevented ad-triggered navigation across all ten apps, requiring Accessibility permission and leaving ads visible roughly one second before dismissal. Documented harms include delayed emergency response, driver distraction, and degraded accessibility for vision-impaired users.

arXiv cs.CR·Sep 15, 15:24 UTC · 1d agoResearch

32

27.5KB language-agnostic WebGPU syntax highlighter

A developer released gpu-lexer, a 27.5KB language-agnostic syntax highlighter that uses a tiny WebGPU model to label code tokens in the browser.

gpu-lexer splits source into words, whitespace, and symbols, then a small WebGPU model uses local and whole-file context to assign nine token classes, working on languages never seen in training. On held-out files, 12.57% of token labels differ from Shiki, though this measures agreement with Shiki rather than objective correctness. In benchmarks against Shiki 4.4.3, Prism.js, Highlight.js, Sugar High, and Starry Night, it highlighted 10 concatenated copies of three.min.js (5.56M characters) about 10x faster on an Apple M4 Pro in Chrome 152. The author frames it as an experiment, not a grammar-equivalent highlighter.

Hacker News · security·Sep 9, 01:14 UTC · 7d agoAI tools & infraHN 21↑ · 5 comments1

15

Attackers plant remote access tools on compromised PaperCut servers

Attackers chained two PaperCut NG/MF zero-days for unauthenticated access, installing SimpleHelp and AnyDesk remote access tools on compromised servers.

An ongoing campaign exploits chained zero-days CVE-2026-81578 (improper access control) and CVE-2026-82078 (unsafe dynamic class loading) in internet-facing PaperCut NG and MF Application Servers, enabling authentication bypass and arbitrary Java bytecode execution. Post-compromise activity includes user and domain enumeration, payload download from sendit.sh, and silent installation of SimpleHelp and AnyDesk for redundant remote access; Defused observed CVE-2026-81578/CVE-2026-82078 exploit activity in honeypots since August 29, including data theft via Derby database dumps. Emergency patches were released August 28 and August 30, but 47% of roughly 2,500 PaperCut installs tracked by Huntress run v23 or older, for which no patch is available.

Help Net Security·Sep 2, 08:12 UTC · 14d agoExploit / PoC in the wildCVE-2026-81578CVE-2026-820781

78

Researchers found a way to hijack devices through Zoom screen sharing

Researchers used a public AI tool to find a Zoom flaw enabling device hijacking via screen sharing in under 20 prompts.

Security researchers discovered a serious vulnerability in Zoom that can be used to hijack devices through the screen-sharing feature. The flaw was reportedly found by a publicly available AI tool in fewer than 20 prompts, highlighting the role of agentic AI in vulnerability discovery. The article does not report active exploitation or assign a CVE in the provided text.

Ars Technica · Security·Aug 12, 13:37 UTC · Aug 12, 2026Vulnerability

65

Context Engineering Inside the Harness: 4 Mechanisms That Beat Context Overflow and Goal Loss on Long-Horizon Tasks

Survey of four harness mechanisms—context budgeting, compaction, todo-state, and memory—that keep long-horizon LLM agents on task across 200+ tool calls.

The article details how agent harnesses, not larger context windows, solve context overflow and goal loss on long-horizon tasks, citing Chroma's Context Rot report showing 18 LLMs (GPT-4.1, Claude 4, Gemini 2.5, Qwen3) degrade on long inputs. Concrete implementations include LangChain Deep Agents offloading tool responses over 20,000 tokens to the filesystem and truncating old tool calls at 85% window usage, and Claude Code capping auto memory at 25KB while re-reading the 5 most recently modified files after compaction. OpenAI's Responses API now offers server-side compaction via context_management with a standalone /responses/compact endpoint, which Codex uses for long-running coding tasks. Manus reports a roughly 100:1 input-to-output token ratio per ~50-tool-call task, motivating todo.md state recitation to prevent goal drift.

MarkTechPost·Sep 13, 05:56 UTC · 3d agoAI research1

52

ZDI-26-602: Foxit PDF Reader Doc Object Use-After-Free Remote Code Execution Vulnerability

ZDI disclosed a use-after-free in Foxit PDF Reader (CVE-2026-13128) enabling remote code execution when a user opens a malicious file.

Zero Day Initiative advisory ZDI-26-602 describes a use-after-free flaw in Foxit PDF Reader's document object handling. Exploitation yields arbitrary code execution but requires the target to open a malicious page or file. ZDI assigned CVSS 7.8 and CVE-2026-13128. The advisory does not report active exploitation.

ZDI Published Advisories·Aug 24, 05:00 UTC · 23d agoVulnerabilityCVE-2026-13128

24

ZDI-26-636: Oracle Outside In Technology PostScript File Parsing Heap-based Buffer Overflow Remote Code Execution Vulnerability

ZDI disclosed a heap-based buffer overflow in Oracle Outside In Technology's PostScript parsing (CVE-2026-60412) enabling user-triggered remote code execution.

Zero Day Initiative published ZDI-26-636, a CVSS 7.8 heap-based buffer overflow in PostScript file parsing within Oracle Outside In Technology. Remote attackers can execute arbitrary code when the target opens a malicious file or visits a malicious page, making user interaction a requirement. The vulnerability is tracked as CVE-2026-60412. The advisory reports no exploitation.

ZDI Published Advisories·Sep 9, 05:00 UTC · 7d agoVulnerability 4 sourcesCVE-2026-604121

22

Zoom Annotation Flaws Could Let a Meeting Participant Hijack Another Attendee's Client

A Security researchers disclosed three Zoom annotation flaws enabling zero-click client hijacking; Zoom shipped fixes in June and July with no exploitation reported.

Researchers at A Security found three flaws in Zoom's annotation feature: CVE-2026-53413 (CVSS 8.3, buffer over-write), CVE-2026-53414 (CVSS 6.5, buffer over-read), and CVE-2026-53415 (CVSS 8.3, use-after-free). A crafted drawing object sent over the wrong message channel can overwrite adjacent memory and hijack another attendee's client with no user interaction. Fixes shipped in Zoom Workplace 7.1.5/7.0.6, VDI Client 7.0.11/6.6.16, and Zoom Rooms/Meeting SDK 7.1.0+ during June and July. No exploitation has been reported and the flaws are absent from CISA's Known Exploited Vulnerabilities catalog.

The Hacker News·Aug 11, 19:08 UTC · Aug 11, 2026VulnerabilityCVE-2026-53413CVE-2026-53414CVE-2026-53415

72

The GNU C Library security advisories update for 2026-09-14

glibc advisory GLIBC-SA-2026-0017 discloses a buffer overflow in strfmon/strfmon_l affecting versions 2.38 through 2.44.

The GNU C Library published security advisories including GLIBC-SA-2026-0017, a buffer overflow in strfmon and strfmon_l. Calling these functions with right-justified width padding conversions can write past the end of the caller-supplied output buffer in glibc 2.38 to 2.44. Exploitation requires an application code path that calls strfmon with attacker-influenced parameters.

oss-security·Sep 14, 19:54 UTC · 1d agoVulnerability 2 sources1

38

ZDI-26-610: Apple Safari JavaScriptCore B3 ReduceStrength Phase Use-After-Free Remote Code Execution Vulnerability

ZDI details a use-after-free in Apple Safari's JavaScriptCore (CVE-2026-64715) that allows remote code execution after a user visits a malicious page.

The Zero Day Initiative published advisory ZDI-26-610 for a use-after-free in the B3 ReduceStrength phase of Apple Safari's JavaScriptCore. Successful exploitation allows remote attackers to execute arbitrary code, but user interaction is required, such as visiting a malicious page or opening a malicious file. ZDI rates the vulnerability 8.8 on CVSS and assigned CVE-2026-64715. The advisory does not report exploitation in the wild.

ZDI Published Advisories·Aug 24, 05:00 UTC · 23d agoVulnerabilityCVE-2026-64715

30

ZDI-26-597: Foxit PDF Reader AcroForm Use-After-Free Remote Code Execution Vulnerability

Foxit PDF Reader has a use-after-free vulnerability (CVE-2026-57252, CVSS 7.8) allowing remote code execution via malicious AcroForm content.

ZDI-26-597 describes a use-after-free vulnerability in the AcroForm feature of Foxit PDF Reader, tracked as CVE-2026-57252 with a CVSS score of 7.8. Successful exploitation allows remote attackers to execute arbitrary code, but requires user interaction such as visiting a malicious page or opening a malicious file. The advisory does not mention any exploitation in the wild.

ZDI Published Advisories·Aug 24, 05:00 UTC · 23d agoVulnerabilityCVE-2026-57252

30

Quantifying the Engagement Trap: Impact of Short-form Video Recommender Systems on Users with ADHD

A 302-participant study finds engagement-optimized short-form video recommenders cause disproportionate time blindness and distress for users with ADHD.

Researchers ran a stratified Prolific study with 302 participants comparing short-form video recommendation experiences with and without ADHD. Participants with ADHD reported significantly higher time blindness, post-usage regret, and emotional distress despite perceiving recommendations as similarly relevant. The paper proposes neurodiversity-aware, human-centered design interventions to mitigate these algorithmic harms.

arXiv cs.AI / cs.LG / cs.CL·Sep 7, 17:38 UTC · 9d agoAI research

18

ZDI-26-600: Foxit PDF Reader Annotation Use-After-Free Information Disclosure Vulnerability

Foxit PDF Reader annotation use-after-free (CVE-2026-57237, CVSS 3.3) can disclose sensitive information when users open malicious PDFs.

ZDI-26-600 documents a use-after-free in Foxit PDF Reader annotation handling that allows remote information disclosure, tracked as CVE-2026-57237 with CVSS 3.3. Exploitation requires user interaction such as opening a malicious file or visiting a malicious page. The advisory was published by the Zero Day Initiative on August 24, 2026.

ZDI Published Advisories·Aug 24, 05:00 UTC · 23d agoAdvisoryCVE-2026-57237

15

ZDI-26-601: Foxit PDF Reader Annotation Use-After-Free Information Disclosure Vulnerability

Foxit PDF Reader has a use-after-free vulnerability (CVE-2026-13129, CVSS 3.3) allowing sensitive information disclosure when a user opens malicious content.

ZDI-26-601 describes a use-after-free vulnerability in the annotation feature of Foxit PDF Reader, tracked as CVE-2026-13129 with a CVSS score of 3.3. Successful exploitation allows remote attackers to disclose sensitive information, but requires user interaction such as visiting a malicious page or opening a malicious file. The advisory does not mention any exploitation in the wild.

ZDI Published Advisories·Aug 24, 05:00 UTC · 23d agoVulnerabilityCVE-2026-13129

15

llm 0.34

Version 0.34 of Simon Willison's llm CLI adds response-duration metrics to log output, plus bug fixes and faster log querying.

The open-source llm command-line tool for interacting with large language models released version 0.34. The headline change adds response duration in milliseconds and human-readable form to llm logs --usage Markdown output, plus a new duration_ms field in llm logs --short. The release includes several contributed bug fixes and a significant performance improvement to llm logs, alongside the related llm-openrouter 0.7.1 release.

Simon Willison·Sep 2, 19:23 UTC · 13d agoAI tools & infra1

28

Over 440,000 Exploit Attempts Target Super Forms and Elementor Pro RCE Flaws

Wordfence blocked 440,000+ exploit attempts against critical unauthenticated RCE flaws in WordPress plugins Super Forms and Elementor Pro.

Wordfence reports mass exploitation of two unauthenticated arbitrary file upload RCE flaws: CVE-2026-14894 in Super Forms (CVSS 9.8, fixed in 6.3.314) and CVE-2026-32475 in Elementor Pro (CVSS 9.0/9.8, fixed in 4.2.2), with over 250,000 and 190,000 blocked exploit attempts respectively. Attackers upload Base64-encoded PHP web shells such as Mushr00w_upl.php to execute code, create administrator accounts, exfiltrate data, or seize sites. Super Forms exploitation began July 14, 2026 and peaked above 40,000 requests on August 18; Elementor Pro attacks started August 19. Successful Elementor Pro exploitation requires a published page with a Form widget containing a File Upload field.

The Hacker News·Sep 4, 08:48 UTC · 12d agoExploit / PoC in the wildCVE-2026-14894CVE-2026-32475

80

ZDI-26-598: Foxit PDF Reader AcroForm Use-After-Free Remote Code Execution Vulnerability

Foxit PDF Reader AcroForm use-after-free (CVE-2026-57242, CVSS 7.8) permits remote code execution through malicious PDF files or pages.

ZDI-26-598 details a use-after-free in the AcroForm component of Foxit PDF Reader, tracked as CVE-2026-57242 and rated CVSS 7.8. A remote attacker can execute arbitrary code if the target opens a malicious file or visits a malicious page. The advisory was published by the Zero Day Initiative on August 24, 2026.

ZDI Published Advisories·Aug 24, 05:00 UTC · 23d agoAdvisoryCVE-2026-57242

25

[remote] PCMan 2.0.7 - Buffer Overflow

A remote buffer overflow in PCMan 2.0.7 has a public proof-of-concept exploit published on Exploit-DB.

Exploit-DB listing 52657 discloses a remote buffer overflow vulnerability in PCMan version 2.0.7. A proof-of-concept exploit is publicly available, allowing attackers to potentially crash or compromise affected instances. The listing includes no CVE identifier and reports no observed exploitation in the wild.

Exploit-DB·Aug 18, 00:00 UTC · 29d agoExploit / PoC1

22

Zoom Patches “Zoomsday” Zero-Click Flaw Enabling Remote Code Execution

Zoom patched CVE-2026-53413, a zero-click annotation flaw dubbed "Zoomsday" allowing remote code execution on meeting participants' devices across all platforms.

Zoom patched four vulnerabilities, including CVE-2026-53413, a stack buffer overflow in CAnnoFormatBlock::Deserialize in the annotation protocol that allows zero-click remote code execution on another participant's device. A Security also found CVE-2026-53414, a buffer overread enabling denial-of-service crashes, and CVE-2026-53415, a use-after-free Zoom had already discovered internally. Updates shipped for Workplace 7.1.5 and 7.0.6, Rooms 7.1.5, and Meeting SDK 7.1.5 across all supported platforms.

Security Affairs·Aug 11, 17:48 UTC · Aug 11, 2026VulnerabilityCVE-2026-53413CVE-2026-53414CVE-2026-53415

70

USN-8736-1: Perl vulnerabilities

Ubuntu issued USN-8736-1 fixing two Perl regex flaws that could cause denial of service, code execution, or security-restriction bypass.

Ubuntu released a security notice addressing two Perl vulnerabilities in regular expression handling. CVE-2026-15534 involves mishandling of large inputs during regex matching, enabling out-of-bounds heap reads or writes that could lead to denial of service or arbitrary code execution. CVE-2026-19487 involves incorrect matching for regexes with alternative branches, allowing security restrictions to be bypassed. Users are advised to apply the updated packages.

Ubuntu Security Notices·Sep 8, 14:25 UTC · 8d agoAdvisoryCVE-2026-15534CVE-2026-19487

25

TLD Tracker: Exploring Newly Released Top

Unit 42 tracked 19 newly released top-level domains and found large-scale phishing, unwanted program distribution, and cybersquatting tied to TLD launch dates.

Researchers analyzed 19 new generic TLDs, including .zip, .bot, .ing, and .meme, released or approaching general availability over roughly 18 months. Data from passive DNS, registry zone files, newly registered domain feeds, and the Tranco top-1M list showed phishing campaigns, potentially unwanted program distribution, and domain squatting on these TLDs. Abuse correlated with each TLD's rollout phases, indicating attackers monitor general availability dates to register and weaponize domains. The IANA root database now lists over 1,000 generic TLDs.

Palo Alto Unit 42·Aug 17, 12:47 UTC · Aug 17, 2026Research

38

LexFlip: A Dissociation Diagnostic for Legal Meaning Preservation Metrics

LexFlip releases 373 minimal perturbations of Quebec statutory French that reverse legal force while preserving tokens, exposing weaknesses in embedding-based meaning preservation metrics.

LexFlip provides 373 minimal perturbations of Quebec statutory French that reverse legal force while preserving 0.93 of tokens, creating dissociation items that break monotone token-overlap metric validation. The seven embedding and BERTScore metrics tested register only 0.022-0.039 of their identical-to-unrelated range on these edits, versus 0.670 for bidirectional NLI. Against FrJudge, with a measured human ceiling of r=0.597, a bare length feature outscores every semantic metric tested.

arXiv cs.AI / cs.LG / cs.CL·Sep 4, 15:49 UTC · 12d agoAI research

22

Unbound: 1.26.1 addresses multiple CVE items

Unbound 1.26.1 fixes critical heap buffer overflow CVE-2026-81642 in DNSKEY digesting that may allow remote code execution.

NLnet Labs released Unbound 1.26.1 on September 16 as a security release fixing several vulnerabilities, most notably CVE-2026-81642, rated CRITICAL, a heap buffer overflow during DNSKEY digesting that can possibly lead to remote code execution. Unbound is a widely deployed DNS recursive resolver and validating resolver. Administrators should upgrade promptly.

oss-security·15:05 UTC · 4h agoVulnerabilityCVE-2026-81642

62