Search: “tamper-localization”

30 stories

DomainAll Cyber AIWindow24h 3d 7d 30d AllIn the wild Important only Show duplicates

CategoryAll

SeverityAll critical 0 high 0 medium 0 low 0 info 0

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

TrajMark: Ownership Attribution and Segment-Level Tamper Localization for Coding-Agent Trajectories

Researchers introduce TrajMark, a training-free watermarking framework for coding-agent trajectories that recovers ownership, detects 95.5-100% of edits, and localizes tampered regions.

TrajMark is a training-free, symmetric-key, visible-only watermarking framework for coding-agent trajectories that separates robust ownership attribution from fragile local integrity verification. A sparse owner layer encodes a six-bit deployment identifier by rewriting keyed READ actions into masked linear equations, while a localization layer inserts linked Q12 seals that commit to protected critical-action segments. Across three coding-agent frameworks and three LLMs, it recovers the exact owner in all clean full-watermark batches, detects 95.5%-100% of single-site edits, and localizes 95.8% of random corruptions to an accepted protocol region. Owner marking adds no trajectory actions and matched Pass@1 is 26.9% versus 26.3% for unwatermarked runs.

arXiv cs.CRSep 9, 16:33 UTC · 6d agoResearch1

NovaFabric: Tamper-Evident, Replayable Evidence for Autonomous AI Agent Runs

NovaFabric seals autonomous AI agent runs into tamper-evident, replayable Run Capsules enabling third-party audit under EU AI Act and ISO 42001.

NovaFabric records autonomous agent runs without modifying agent logic into portable Run Capsules (fifteen-entity schema) sealed with DSSE signatures, RFC 3161 timestamps, a Merkle log, and redaction attestations, supporting four-mode replay and third-party Evidence Bundle verification. Evaluation shows tampering rejected across three tested classes, 14/14 credential types redacted while preserving 9/9 decoys, and 140/140 mutations localized; blast-radius queries reach 45.5ms p99 over 10M edges. Limits include only 2/10 tool-using workloads completing replay due to missing tool-response substitution and ingest capped at 61.6 req/s. The contribution integrates OpenTelemetry, DSSE/in-toto, and W3C PROV rather than new cryptography.

arXiv cs.CRSep 11, 08:35 UTC · 5d agoAI safety & security1

Vulnerability Localization Benchmark: Measuring Agentic Security Analysis at Repository Scale

VLoc Bench tests 27 language models at locating vulnerable files in 290 repositories; best system reaches 0.229 File F1 and 38.4% of tasks unsolved.

The Vulnerability Localization Benchmark (VLoc Bench) contains 500 real-world vulnerabilities from 290 repositories across six package ecosystems and 147 CWE categories, pairing pre-fix and post-fix repository snapshots. Agents receive only a CWE description and read-only terminal access to identify affected files, and must confirm absence on patched snapshots. The strongest of 27 language models and four static-analysis tools achieves just 0.229 File F1; 38.4% of tasks receive no correct localization, and effective localizers still report unsupported locations on patched repositories.

arXiv cs.AI / cs.LG / cs.CLSep 14, 17:44 UTC · 1d agoAI research

Attackers conceal phishing lures using invisible Unicode characters

Threat actors use invisible Unicode characters (ASCII smuggling) to hide phishing lures and evade email security filters.

Threat actors have adopted the ASCII smuggling technique in phishing campaigns, embedding invisible Unicode characters in emails to conceal malicious lures. The approach is designed to evade email security filters that scan for visible phishing indicators. The report gives no victim counts or named campaigns.

BleepingComputerSep 6, 14:23 UTC · 9d agoPhishing & fraud in the wild

LLM Forensics: Where Do Backdoors Hide? Localizing and Controlling Trigger Mechanisms with Sparse Autoencoders

Researchers use sparse autoencoders to localize trigger-based backdoor mechanisms in 1B and 8B LLMs, finding detection features differ from causal control features.

In a controlled language-switching backdoor setting where fixed trigger sequences make 1B and 8B language models continue English prompts in French or German, the authors train sparse autoencoders (SAEs) across layers and transformer components. Attention and MLP features detect triggered prompts with near-perfect F1, but ablating them rarely suppresses the language switch, while residual-stream features can suppress triggered generation and some can induce target-language continuations without the trigger. The work decomposes token-trigger mechanisms into distinct SAE feature roles: trigger detection, residual-stream propagation, and language tracking, a decomposition the authors expect to transfer to other trigger-based backdoors.

arXiv cs.AI / cs.LG / cs.CLSep 7, 16:47 UTC · 8d agoAI safety & security1

Once popular for attacking AI, ASCII smuggling is embraced by spammers

Spammers adopt ASCII smuggling—invisible Unicode tag characters—to evade email filters, with Microsoft Defender detections spiking to 2.5 million per day.

ASCII smuggling hides text in Unicode tag characters (e.g., U+E0041 for "A") that are invisible to humans but readable by LLMs and text processors. The technique gained attention as a stealthy prompt-injection vector and is now used by spammers to obfuscate keywords from email detectors. Microsoft reported Defender for Office smuggling detections jumped from roughly 21,000 per day to over 1.3 million in early February, reaching 2.5 million within four days, before falling sharply in mid-May.

Ars Technica · SecuritySep 4, 17:18 UTC · 11d agoPhishing & fraud

The Crypto Wallet That Never Opened: Tampered Exodus Installer Hides a Modular RAT

Huntress found tampered Exodus crypto wallet installers delivering a modular RAT that steals credentials rather than wallet funds.

Huntress analysts analyzed tampered installers for the Exodus cryptocurrency wallet that bundle a modular remote access trojan. The implant focuses on harvesting credentials instead of draining wallet balances, suggesting broader access theft. The case highlights installer tampering as a supply-chain-style delivery vector for credential-stealing tooling.

HuntressSep 1, 04:00 UTC · 15d agoMalware in the wild

AsyncRAT Malware Abuses AutoIt and PowerShell to Hide Inside Legitimate Windows Process

Point Wild details a five-stage AsyncRAT campaign using AutoIt abuse and process injection to hide a .NET RAT inside Microsoft's signed charmap.exe.

Researchers at Point Wild Threat Intelligence analyzed a campaign starting from a lure batch file ('Right-click to open Invoice Details.bat') that launches hidden PowerShell to reconstruct a Base64/XOR-obfuscated payload. The stage drops a renamed signed AutoIt interpreter and loader script into %LOCALAPPDATA%\Temp and achieves persistence via a batch file in the user's Startup folder (T1547.001). The loader decrypts the payload in memory with single-byte XOR key 0x36 and injects it into SysWOW64\charmap.exe using OpenProcess/VirtualAllocEx/WriteProcessMemory/CreateRemoteThread, with PE-sieve confirming an in-memory implanted PE and patched AMSI modules. The final AsyncRAT payload performs screen capture and data theft, with observed C2 at 158[.]51[.]122[.]136:4944 over raw TCP.

GBHackersSep 14, 05:24 UTC · 2d agoMalware3· 1 read

Understanding the Usability of Cryptographic Verification Tools

Survey of Tamarin and ProVerif users reveals usability barriers: debugging non-termination, model validation, and opaque proof failures hinder cryptographic protocol verification.

The paper presents an exploratory human-centered survey of researchers, graduate students, and practitioners with hands-on experience using Tamarin, ProVerif, and related cryptographic protocol verification tools. Findings reveal usability barriers across the verification workflow, including difficulties debugging non-termination and performance issues plus the lack of systematic methods for validating formal models against real protocols. When proofs fail without concrete attacks, users commonly simplify models, add helper lemmas, and revisit modeling abstractions. Participants called for actionable diagnostics, clearer explanations of results, visualization, and automation for recurring proof tasks.

arXiv cs.CRSep 14, 20:40 UTC · 1d agoResearch

ZK-Trace: Certified Collusion Tracing with Zero-Knowledge Credentials for Federated GNSS Interference Monitoring

ZK-Trace combines Tardos fingerprints and zero-knowledge credentials to trace leaked classifier copies in federated GNSS monitoring without leaker cooperation.

ZK-Trace addresses leakage of a proprietary classifier distributed to partly trusted stations in federated GNSS interference monitoring, combining public identity marks, recipient-specific Tardos fingerprints, and zero-knowledge credential verification to support offline tracing. The paper provides false-accusation and tracing-score bounds with an interval-arithmetic checker allocating a common budget across accusation and tamper decisions. In a simulated GNSS federation it isolates all 160 single-owner copies and traces 712 of 720 two-owner mixtures with a 0.001 false-naming budget, while feature marks survive feature matching in 20/20 runs at 4.8 percentage-point accuracy cost but are erased by function-only distillation.

arXiv cs.CRSep 8, 14:00 UTC · 7d agoResearch

ActGuard: Pre-execution Action Auditing against Indirect Prompt Injection in LLM Agents

ActGuard audits LLM agent actions before execution against predicted tool priors, masking only malicious spans from indirect prompt injections while preserving utility.

ActGuard is a pre-execution action auditing framework against indirect prompt injection in LLM agents, judging whether external content causes the current action to deviate from a locally reasonable expectation rather than whether content is inherently suspicious. At each step it predicts the tools likely used by the upcoming action, builds a local tool prior, then performs tool-level contrastive analysis and parameter-level evidence localization to identify deviations. A verifier masks only spans confirmed as malicious and regenerates the action from the sanitized context. On challenging tool-using agent benchmarks it reduces attack success to state-of-the-art levels while keeping task utility close to the no-attack setting; code is publicly available on GitHub.

arXiv cs.CRSep 14, 03:47 UTC · 2d agoAI safety & security

CVE-2026-84939: Apache FreeMarker: A malformed locale may be exploitable for path traversal attacks

Apache FreeMarker CVE-2026-84939 permits path traversal via a malformed locale in template loading; fixed in FreeMarker 2.3.35.

A path traversal vulnerability in Apache FreeMarker's template loading mechanism can be triggered when an attacker controls a malformed locale value. FreeMarker 2.2.0 through 2.3.34 are affected for both the org.freemarker:freemarker and freemarker-gae artifacts, while 2.3.35 is unaffected. No exploitation or CVSS details were included in the announcement.

oss-securitySep 8, 15:27 UTC · 7d agoVulnerability CVE-2026-84939

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.

Ars Technica · SecurityAug 20, 13:00 UTC · 26d agoAI safety & security1

Cross-Domain Inference for Human Localization: Applying Wi-Fi RSSI Data to CSI-Trained Models

Researchers show CSI-trained Wi-Fi models can localize people from RSSI data at ~80% confidence, enabling privacy attacks from ordinary IoT devices.

The paper investigates cross-domain inference, feeding RSSI data into an existing CSI-based Wi-Fi pose prediction model. RSSI is accessible on IoT devices without elevated OS permissions or specialized drivers, unlike CSI. Using an RSSI dataset synchronized with video ground truth, the model predicted human locations with approximately 80% confidence when movement was present. The results imply a wide range of commodity IoT devices could be used for privacy invasion in Wi-Fi-dense environments.

arXiv cs.CRSep 15, 13:56 UTC · 21h agoResearch

[webapps] Langflow 1.8.4 - Path Traversal to Remote Code Execution

A path traversal to remote code execution exploit for Langflow 1.8.4, a popular LLM application builder, was published on Exploit-DB.

Exploit-DB lists a proof-of-concept exploit chaining path traversal to remote code execution in Langflow 1.8.4, an open-source tool used to build LLM applications and agents. The chain allows an attacker to write arbitrary files outside the intended directory and achieve code execution on the host. The provided text does not include a CVE identifier or reports of exploitation in the wild, but RCE in a widely deployed AI tooling product is notable for defenders.

Exploit-DBAug 31, 00:00 UTC · 16d agoExploit / PoC1

Linux Detection Engineering - Local Privilege Escalation

Elastic details a layered detection framework for Linux local privilege escalation, covering 2026's copy-on-write bug wave and LLM-assisted discovery.

Elastic Security Labs describes how most Linux local privilege escalations share a common host flow — an unprivileged process launched from a writable path becoming root — and proposes layered detections combining general outcome-based rules with per-technique rules in Elastic Defend and Auditd. It tracks 13 recent LPE disclosures, seven of which share a copy-on-write/zero-copy bug class, including Copy Fail, DirtyFrag, Fragnesia, DirtyDecrypt, DirtyClone, pedit COW, and RefluXFS. Qualys attributes RefluXFS to an LLM-assisted research effort with Anthropic using Claude Mythos Preview, and another bug is credited to an LLM-assisted workflow. Detection and endpoint rules are published in Elastic's detection-rules and protections-artifacts repositories.

Elastic Security LabsSep 11, 00:00 UTC · 5d agoResearch

Attackers Exploit Critical Langflow and Rails Flaws in Credential

VulnCheck reports active exploitation of critical Langflow CVE-2026-0768 and Rails CVE-2026-66066 for credential harvesting, with detections rising to 360.

VulnCheck observed active exploitation of CVE-2026-0768 (CVSS 9.8) in Langflow and CVE-2026-66066 'KindaRails2Shell' (CVSS 9.5) in Ruby on Rails, with detections rising from 50 on August 30, 2026 to 360 by September 1. The Rails flaw allows unauthenticated arbitrary file reads, leaking secret_key_base, Rails master key, database passwords, cloud credentials and API tokens, ultimately enabling RCE; the patch still leaves the variation-key Marshal deserialization RCE gadget functional. Observed chains include a Python credential harvester with SimpleHelp remote access via CVE-2026-5027, and weaponization of CVE-2025-3248 to enlist hosts into an XMR mining botnet after disabling auditd. More than 7,100 exposed vulnerable Ruby on Rails instances and over 15,000 successful exploitation attempts across three Langflow flaws were recorded.

The Hacker NewsSep 1, 07:23 UTC · 15d agoExploit / PoC in the wildCVE-2026-0768 CVE-2026-66066 CVE-2026-0769+2 CVEs1

Tricky 'SynkLoader' Multitool May Herald Ransomware

Researchers detail SynkLoader, an advanced multilingual multitool that hijacks screens to steal passwords and may precede ransomware attacks.

Dark Reading reports analysis of SynkLoader, an advanced, multilingual malware family functioning as a multitool loader. The family revives an older technique — screen hijacking — to enable effective password theft, alongside several novel features. Analysts assess its activity may be a precursor to ransomware deployments, making early detection valuable for defenders.

Dark ReadingAug 24, 15:02 UTC · 22d agoMalware

Threats Making WAVs - Incident Response to a Cryptomining Attack

Guardicore researchers dissect a cryptomining attack that hid a cryptominer inside WAV files, mapping the full infection chain and response steps.

Guardicore security researchers present a full analysis of a cryptomining attack that concealed a cryptominer inside WAV audio files. The report documents the complete attack chain from detection through infection, network propagation, and malware analysis. It also includes recommendations for optimizing incident response processes in data centers.

Akamai BlogSep 8, 15:13 UTC · 7d agoMalware in the wild

ClickFix moves into the browser: Cryptocurrency theft with Google-hosted C2

Talos tracks a crypto-theft campaign abusing the Google Visualization API for C2, using browser-based ClickFix lures to inject web-skimmer JavaScript.

Criminal actors convinced targets to paste JavaScript into Chrome's address bar or install it in the Tampermonkey extension, injecting a web skimmer into sessions on two cryptocurrency trading websites; Tampermonkey also provides persistence. Since March 2026 the campaign retrieved obfuscated scripts via the Google Visualization API from public Google Sheets documents, hooking the fetch API and replacing cryptocurrency deposit addresses in responses and the clipboard. Lures pose as leaked reports of a nonexistent API flaw at cryptocurrency swap services and spread via Telegram, DarkForums, and paste sites since early October 2025. Talos warns the legitimate-service-abuse techniques could enable wider supply-chain attacks on e-commerce and customer-facing systems.

Cisco TalosSep 8, 10:00 UTC · 8d agoPhishing & fraud in the wild

ASCII smuggling crosses over from AI prompt injection to phishing evasion

Microsoft details high-volume phishing campaign using ASCII smuggling (Unicode tag chars) for filter evasion, peaking at 2.3M messages.

Microsoft researchers observed a high-volume finance-themed phishing campaign using invisible Unicode tag characters (U+E0000–U+E007F), a technique known from AI prompt injection research as ASCII smuggling, to split lure words like 'funding' and evade email filters. Telemetry from Microsoft Defender for Office 365 showed signature hits jump from roughly 21,000 messages on February 8, 2026 to more than 1.3 million on February 9, peaking above 2.3 million on February 11, with elevated weekday activity lasting approximately three months. The discovery emerged from prompt injection protection research, showing AI-era evasion techniques crossing into traditional phishing. Most messages were flagged by layered Defender protections rather than a single Unicode-specific signal.

Microsoft Security BlogSep 3, 16:00 UTC · 12d agoPhishing & fraud in the wild

Hackers Can Hide Malicious AI Commands Inside Normal English to Bypass Security Filters

Check Point's PuzzleMask technique hides malicious prompts in ordinary English that fast gatekeeper models miss but high-reasoning downstream models execute.

Check Point researchers disclosed PuzzleMask, a technique concealing policy-breaking instructions in natural-language prose without encodings or invisible characters. Fast screening models classified all 23 crafted wrappers as safe, while a high-reasoning model recovered and acted on the hidden instruction in 17 of 18 tests (94.4%). The gap stems from capability imbalance between gatekeeper and target models, with defenses including paraphrasing untrusted input, stricter self-referential wording rules, and output/tool-call monitoring.

Cyber Security NewsSep 11, 07:36 UTC · 5d agoAI safety & security1

When the prompt becomes the payload: A practical pen-testing guide for GenAI, LLM and RAG applications

CSO Online publishes a practical penetration-testing guide for GenAI, LLM, and RAG applications, covering prompt injection, retrieval poisoning, and tenant isolation testing.

The guide frames LLM applications as attack graphs spanning prompts, retrieval layers, vector stores, tools, identities, and downstream APIs, arguing that conventional web testing misses instruction-vs-data channel risks. It builds on OWASP prompt injection guidance (direct vs. indirect injection) and NIST's 2025 adversarial machine-learning taxonomy, noting that RAG and fine-tuning do not remove injection risk. Recommended practices include documenting trust transitions across components, using canaries and synthetic records to avoid test side effects, running multi-turn and obfuscated injection campaigns, and verifying chains from poisoned documents to observable state changes. It also details testing RAG pipelines via controlled document poisoning across metadata, OCR layers, and code comments, plus cross-tenant isolation checks on retrieved document IDs.

CSO OnlineSep 9, 10:00 UTC · 7d agoAI safety & security1

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.

GBHackersSep 11, 05:53 UTC · 5d agoAI safety & security 2 sources

Hackers Abuse AutoIt to Inject AsyncRAT Into Microsoft-Signed Windows Process

Attackers use AutoIt and an invoice-themed batch file to inject AsyncRAT into Microsoft-signed charmap.exe for stealthy remote access.

Point Wild Threat Intelligence documented a five-stage campaign starting with 'Right-click to open Invoice Details.bat' that launches hidden PowerShell, drops a renamed AutoIt interpreter and encrypted loader into a random Temp folder, and persists via a Startup-folder batch file. The loader decrypts AsyncRAT only in memory and injects it into the Microsoft-signed charmap.exe (Windows Character Map) process, leaving no payload file on disk. The final payload patches AMSI inside the infected process and includes screen-capture code that prepares images for transfer to a command-and-control server at 158[.]51[.]122[.]136:4944.

Cyber Security NewsSep 14, 07:07 UTC · 2d agoMalware in the wild1

Enoki: Efficient Multi-Level Hallucination Detection

Researchers introduce Enoki, an open information extraction framework unifying claim-level and span-level hallucination detection in LLMs at lower inference cost.

Enoki extracts text-anchored relational facts, verifies them against evidence, and projects unsupported facts back onto hallucinated spans, so claim-level verification and span-level localization share one representation without separate alignment. It supports LLM-based, encoder-based, and rule-based extraction regimes to balance accuracy against inference cost. Experiments show it stays competitive with strong claim-level systems while using fewer resources and outperforms them on fine-grained span- and entity-level localization. The authors also release EnokiQA, a dual-granularity dataset with aligned claim-level verification and span-level localization annotations.

Hugging Face daily papersAug 31, 22:21 UTC · 15d agoAI research

What Breaks Local Watermarks? A Robustness Benchmark for Local Invisible Image Watermarking

First systematic robustness benchmark of five local invisible image watermarking methods across 55 transformations finds all are vulnerable, with inpainting and geometric misalignment completely breaking payload…

The paper presents the first systematic robustness benchmark for local invisible image watermarks, covering 55 image transformations across signal distortions, coordinate alignment changes, indirect local edits, and direct watermark edits. It evaluates five methods: MaskWM, WAM, OmniGuard, TrustMark, and PixelSeal, all supporting localization natively or with minimal adaptation. Results show every method is vulnerable to some transformation; MaskWM offers the strongest payload recovery and localization but the lowest clean-image quality, and synchronization further improves its recovery under geometric transformations. Geometric misalignment and generative local edits such as inpainting and outpainting can completely impair payload recovery, while signal distortions are often tolerated.

arXiv cs.CRSep 15, 08:30 UTC · 1d agoResearch

LLM-Based Penetration Testing in the Presence of Honeypots

Studies honeypot-aware budget allocation for LLM attack agents, showing detector-guided policies let agents skip deception and compromise real hosts efficiently.

The paper formalizes LLM attacker behavior against honeypots as a budgeted decision process, where agents choose to continue or skip targets when honeypot suspicion arises. A detector-guided policy lets LLM agents allocate execution budget effectively across a mixed host pool in a controlled testbed. Findings show LLM-driven attackers can reason about heterogeneous artifacts and use honeypot suspicion to guide target selection, challenging traditional deception defenses that rely on realism and obscurity against human or script-driven attackers.

arXiv cs.CRSep 8, 01:09 UTC · 8d agoResearch

TasmScan: Continuation-Aware Taint Analysis for TVM Bytecode with Savelist Abstraction

TasmScan introduces source-free taint analysis for TON smart-contract bytecode, detecting 95.3% of defects with 96.8% precision and 17x speedup.

TasmScan is the first bytecode-level static analysis framework for the TON Virtual Machine, enabling cross-continuation data flow reasoning without source code by modeling savelist semantics through forward register analysis with formal over-approximation guarantees. It lifts bytecode into a typed intermediate representation (TASIR) and performs path-sensitive taint analysis. On a 208-contract benchmark with human-confirmed ground truth it detects 95.3% of defects across five classes at 96.8% precision, and resolves 294,546 dynamic continuation targets with 100% precision across 2,921 registry contracts. It achieves a 17x median speedup over symbolic-execution baselines.

arXiv cs.CRSep 15, 11:01 UTC · 1d agoResearch

BenchShield: Formal Model-Backed Instrumentation for Reward Integrity in LLM-Agent Evaluation Infrastructure

BenchShield uses lifecycle-model-backed instrumentation to detect reward hacking in LLM-agent benchmarks, lifting full-chain recall to 77-100% at up to 65% lower cost.

The framework grounds reward-hacking detection in a finite lifecycle model of an evaluation's reward-relevant events, combining a static phase-aware taint analysis with runtime infrastructure-side evidence attribution. Evaluation used a human-labeled corpus of 456 adjudicated trajectories drawn from more than 31,000 public agent runs across three benchmarks. BenchShield improves full-chain recall from 23-94% to 77-100% and same-vector coverage from 16-56% to 43-78%, cuts per-task cost by up to 65%, and achieves 96% accuracy detecting reward hacking at runtime.

arXiv cs.CRSep 10, 03:10 UTC · 6d agoAI safety & security1