Search: “Salesloft Drift”

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

McKesson copes with fallout from data theft extortion attack

McKesson discloses a data theft extortion attack by ShinyHunters affecting oncology and medical-surgical customers, with a reported $55 million demand.

McKesson disclosed that attackers gained access to some of its third-party applications and stole data associated with a subset of customers in its oncology, multispecialty, and medical-surgical business units; the intrusion ran for four days from August 21 and was discovered August 25. ShinyHunters claimed responsibility and listed McKesson on its data-leak site, reportedly demanding more than $55 million with a September 1 deadline. Flashpoint analysts say the group typically uses social engineering and identity weaknesses with valid credentials to access cloud-hosted environments, making the intrusion hard to detect. McKesson, which distributes about one-third of pharmaceuticals used in North America with $403.4 billion in annual revenue, says operations continue and it has reasonable assurance of no ongoing unauthorized activity.

CyberScoop·Aug 31, 21:39 UTC · 16d agoData breach in the wild

80

Navigating the Latent Manifold: Proactive Concept Drift Adaptation for Resilient NIDS

Researchers propose DriftXpert, a concept-drift-adaptive network intrusion detection system validated on enterprise networks, addressing degraded AI-based NIDS performance in dynamic traffic.

AI-based network intrusion detection systems assume static data distributions and degrade under concept drift, raising false positives in dynamic environments. DriftXpert uses a two-stage offline framework: an unsupervised latent-manifold anomaly metric to detect traffic drift, and representation consistency alignment with cross-epoch neuron weight aggregation and selective freezing to transfer knowledge without catastrophic forgetting. Experiments on public datasets and a real enterprise network show effective adaptation to drifted data while retaining known-attack detection.

arXiv cs.CR·Sep 8, 11:58 UTC · 8d agoResearch

36

Mathspace discloses data breach affecting over 1 million people

Mathspace disclosed a Metabase breach exposing data of 1,079,819 students, parents, and staff in Australia and New Zealand.

Mathspace confirmed attackers exploited a vulnerability in its self-hosted Metabase reporting system, gaining administrator access without legitimate login and downloading data on over 1 million people (1,079,819 total) in Australia and New Zealand. Access began August 10, data was downloaded August 27, and the theft was confirmed September 3, 2026. No credentials, academic records, or school-account links were exposed, but affected individuals are warned of targeted phishing. The incident joins a broader campaign against Metabase instances, including Trezor's provider ShipMonk, Framework, and Tally, linked to ShinyHunters via extortion emails and leak-site listings.

BleepingComputer·Sep 7, 13:05 UTC · 9d agoData breach in the wild

78

DRIFT: Removing Diffusion Watermarks by Deflecting the Generative Trajectory

Introduces DRIFT, a black-box attack removing diffusion watermarks by deflecting generative trajectories, achieving 98-100% success across nine watermarking schemes.

Researchers propose DRIFT, a black-box watermark removal attack combining partial forward diffusion with stochastic reverse resampling to break trajectory-dependent verification. The paper derives information-theoretic and Wasserstein source-dependence bounds and shows the first verifier-rejected rung is least distorted among rejected rungs. Across nine watermarks spanning three paradigms, DRIFT achieves 98-100% attack success with the best image quality among compared attacks, without secret keys, verifier internals, or per-image gradient optimization.

arXiv cs.CR·Sep 8, 03:54 UTC · 9d agoResearch

38

What LLM Trading Agents Actually Do in Production: A Six-Month, Population-Scale Record from Two Fleets

Six-month record of 7.5M LLM trading agent invocations shows volatility-blind sizing, minimal upside capture, and no directional edge across two fleets.

The study records autonomous LLM trading agents in production across DX Terminal Pro (3,505 user-funded vaults trading real ETH in Base memecoin markets) and the DXAP fleet (500-599 agents on Hyperliquid perpetuals), spanning roughly six months, 7.5M single-model invocations and about 300K onchain actions. A risk slider explains leverage (+0.425 per level), median leverage is 5.0x in every volatility sextile, and one posture-slider cell holds 62% of liquidations. Agents capture little upside: 43.2% of positions saw +300 bps favorable excursion within 24h yet 49.3% of those closed negative, while the DXAP fleet trails a matched retail benchmark (41% vs 50% roundtrip win rate). A paired-replay league of frontier models finds decision quality statistically indistinguishable at this horizon.

Hugging Face daily papers·Sep 3, 20:00 UTC · 13d agoAI research

30

Show HN: Sunk Cost – How long until a local LLM rig pays for itself?

Show HN tool 'Sunk Cost' calculates when a local LLM rig breaks even versus falling API prices, factoring electricity cost and inference speed.

A Hacker News Show HN project called Sunk Cost models the payback period of buying local LLM hardware instead of paying API prices. Users can adjust assumptions like electricity cost ($/kWh) and API speed (tokens/second), and the model assumes API prices keep falling. Where local speed is unmeasured, it is estimated from memory bandwidth divided by bytes read per token, and labelled as an estimate.

Hacker News · security·Sep 15, 01:37 UTC · 2d agoAI tools & infraHN 30↑ · 31 comments

12

EVOHARNESSBENCH: Can Your Agents Keep Pace with an Evolving Harness?

Researchers introduce EVOHARNESSBENCH, a benchmark showing that evolving agent harnesses (tools, skills, agents) cause forgetting and inconsistent adaptation across 802 tasks.

The paper introduces EVOHARNESSBENCH, a benchmark that places non-stationarity in the externally supplied agent harness rather than in the task stream, evaluating agents across tools, skills, and specialist agents. It comprises 17 multi-stage harness streams built deterministically from verifier-based benchmarks, totaling 802 tasks, 520 tools, 42 skills, and 62 agents. Evaluation covers deployment (retention of previously accessible competence) and self-evolving adaptation settings. Results show harness expansion alone degrades previously solved tasks (harness-induced forgetting), adaptation gains are inconsistent, and retention and adaptation can pull in opposite directions.

Hugging Face daily papers·Sep 2, 20:00 UTC · 14d agoAI research

40

Crypto customers targeted by scammers after email marketing provider breach

Attackers exploited a Brevo SAML SSO flaw to access 138 accounts and phish crypto customers of Trezor, CoinTracking, and BitBox.

An attacker exploited a flaw in Brevo's SAML SSO handling to access 138 customer accounts on September 10; six accounts were used to send phishing emails and contacts were exported from 43 accounts. Crypto firms Trezor, CoinTracking, and BitBox confirmed customers received phishing emails, with Trezor warning roughly 347,000 newsletter subscribers. The Trezor-themed email cited a fabricated STM32 microcontroller entropy bug and urged recipients to enter wallet backups through a malicious link. Exported contact lists could fuel future targeted phishing attacks.

Malwarebytes Labs·Sep 11, 15:01 UTC · 5d agoData breach in the wild 5 sources

75

Cadence: Error-Bounded Lossy Compression of Demand Time Series with a Time-Series Foundation Model

Cadence pairs Google's 330M-parameter TimesFM-3 foundation model with adaptive arithmetic coding, gaining 13-28% on 2026 demand series over classical predictors.

Cadence is an error-bounded lossy compressor for numeric time series combining the 330M-parameter Google TimesFM-3 foundation model with an adaptive arithmetic coder, guaranteeing a per-sample error bound. On 49 EIA-930 balancing-authority demand series from 2026 it gains 13.3% over the best of six classical predictors and 28.3% on 50 MTA ridership series, winning all 297 series-tolerance pairs with a 21.4% median gain. The paper also reports negative results, including that foundation models add negligible value for lossless coding and that PyTorch predictions are not bit-identical across batch sizes.

Hugging Face daily papers·Sep 4, 20:00 UTC · 12d agoAI research1

25

When LLM Decompilers Recompile More and Preserve Less

Researchers show LLM decompiler outputs can recompile yet diverge behaviorally, proposing the Decompile-Diverge fuzzing oracle to catch hidden changes.

The paper demonstrates that LLM-based decompilers can produce code that recompiles and passes all shipped tests yet diverges on other legitimate inputs—4.9% overall and up to 13% for one system—and can make disclosed vulnerabilities vanish without a visible crash. Across 300 real GitHub functions and 287 CVE-grounded functions, a refinement LLM lifted Ghidra's build rate from 75% to 90% while Matched rate fell from 74% to 62%, with up to one tenth of vulnerabilities showing Crash Absence. Decompile-Diverge detects these gaps by synthesizing drivers, growing fuzzing corpora from the reference, and rerunning decompiled code on identical inputs.

arXiv cs.CR·Sep 4, 17:16 UTC · 12d agoResearch

42

Root-Cause Attribution Is a Search Problem: Continual Search for Long-Horizon Agent Failures

Continual Search framework iteratively prompts LLM judges to keep searching agent execution logs, boosting long-horizon failure root-cause attribution accuracy.

The paper frames automated root-cause attribution (RCA) for long-horizon AI agent failures as a search problem, since relevant evidence is sparse and distributed across massive execution traces. The authors propose Continual Search, an iterative framework that nudges an LLM judge across successive turns to keep hunting unresolved diagnostic evidence instead of settling on an early plausible diagnosis. They introduce MegaRCA-Mix, a benchmark of 50 human-annotated failure trials on long-horizon, execution-heavy tasks. On MegaRCA-Mix, Continual Search improves GPT-5.5's F1 from 0.349 to 0.498 (over 40% gain), and lower-tier models can surpass higher-tier counterparts when search is effective.

Hugging Face daily papers·Sep 10, 20:00 UTC · 6d agoAI research1

42

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 Talos·Sep 8, 10:00 UTC · 8d agoPhishing & fraud in the wild

48

From 88 lines to 1: Detecting DLL hijacking with Elastic Defend

Elastic condenses 88 lines of detection logic into one Elastic Defend 9.5.0 rule catching the DLL hijacking used by a ClickFix campaign.

The ClickFix campaign side-loads a malicious mscoree.dll next to a signed Microsoft binary and also ships a driver to kill Elastic Endpoint. Elastic rebuilt the DLL as a NativeAOT library, dropped it beside the signed binary, and confirmed that Elastic Defend 9.5.0 flags the load. The post replaces 88 lines of detection logic with a single rule for DLL search-order hijacking.

Elastic Security Labs·Aug 28, 00:00 UTC · 20d agoTools in the wild

28

StochBench: A Domain-Specific Benchmark for Stochastic Processes in Lean

StochBench introduces 450 graduate-level stochastic-processes problems in Lean 4; an Opus 4.8-based agent proved 34.9% under a 15-minute limit.

StochBench is a Lean 4 benchmark of 450 graduate stochastic-processes problems, each paired with its natural-language source, covering Markov chains, renewal processes, martingales, Brownian motion, stochastic calculus, weak convergence, and Poisson and continuous-time Markov processes. The benchmark addresses field-specific applied mathematics underrepresented in Mathlib, unlike competition-math-dominated suites such as IMO and Putnam collections. An Opus 4.8-based agent achieved a 34.9% proof rate (157/450) under a 15-minute per-problem limit, showing the benchmark remains challenging for advanced provers.

Hugging Face daily papers·Sep 7, 20:00 UTC · 9d agoAI research

25

The Frontier AEO Tracker: What Astra Chooses (and every other frontier model, and what you can do about it)

Latent Space launches an AEO tracker scoring 7 frontier models' product recommendations across 161 categories, revealing generational bias flips.

Latent Space built a tracker measuring Answer Engine Optimization by running 6 prompt variations across 7 frontier models with search enabled over 161 product categories, scoring first choices, alternatives, mentions, and anti-recommendations. It found 28 categories with a universally dominant primary choice and observed soft biases, such as models favoring their own lab's coding agents. Analysis of Anthropic's Sol→Astra and Opus→Fable generations showed newer models consulting fewer sources and being less likely to change answers when questions are paraphrased.

Latent Space·Sep 7, 21:32 UTC · 9d agoAI research

35

EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agentsnew

EvolveTrade lets LLM trading agents self-refine their tool-use policy from realized portfolio feedback, improving Sharpe ratios.

EvolveTrade treats a tool-using trading agent's system prompt as a text-parameterized policy that a Policy Agent revises after each update interval using accumulated decision traces and realized portfolio feedback, keeping the backbone LLM fixed. Experiments across multiple market regimes and two LLM backbones show improved Sharpe Ratio and Cumulative Return over fixed-policy baselines in most settings. Behavioral analyses show evolved policies increase code-mediated analysis and activate regime-relevant computations, with case-level attributions linking policy changes to returns.

Hugging Face daily papers·Sep 14, 20:00 UTC · 2d agoAI research

35

General Quantification of Covariate and Concept Shifts

Paper proposes γ*-concept shifts via entropic optimal transport, deriving estimable generalization bounds unifying covariate and concept shift under distribution shift.

The authors show existing definitions of concept shift break when source and target supports mismatch and propose γ*-concept shifts grounded in entropic optimal transport. They derive a general error bound covering broad loss functions, label spaces and stochastic labeling, plus estimators with concentration guarantees. The resulting DataShifts algorithm quantifies distribution shifts and estimates the error bound in most applications, addressing learning bounds that were previously non-estimable from samples.

arXiv cs.AI / cs.LG / cs.CL·Sep 10, 17:57 UTC · 6d agoAI research1

28

The 12 Best Managed Detection & Response (MDR) Services, Compared and Priced

Buyer's guide compares 12 MDR services, naming Huntress best value, CrowdStrike Falcon Complete for response authority and Expel for transparency.

The article compares 12 managed detection and response providers across response authority, tool bundling and pricing, highlighting Huntress for published SMB pricing and CrowdStrike Falcon Complete for unilateral containment. It stresses the consolidation landscape: Sophos completed its acquisition of Secureworks in February 2025 for approximately $859 million, and Arctic Wolf closed its purchase of BlackBerry's Cylance endpoint assets the same month. It also warns that only full-response contract tiers isolate hosts and kill processes, while lower tiers only triage or guide.

GBHackers·updated Sep 9, 08:07 UTC · 7d ago·first Sep 9, 07:56 UTC · 7d agoIndustry 3 sources1

15

The Evolution of the Agent Harness

Latent Space essay argues late-2025 agent gains came from models and harnesses maturing together, with harness logic absorbed into model weights.

The piece defines the agent harness as everything beyond model weights—tools, context, memory, guardrails—and charts its evolution from ReAct prompting (October 2022) through AutoGPT's premature autonomy, Cursor/Copilot's human-in-the-loop retreat, and Devin's roughly 15% success rate, to o1's capability overhang and Claude Code's February 2025 terminal agent with permission rules. It argues the Christmas 2025 jump cited by Transformer co-inventor Lukasz Kaiser reflected model and harness curves crossing, and that remaining harnesses will serve human attention rather than the model.

Latent Space·Aug 22, 07:30 UTC · 25d agoAI tools & infra

32

Passkey-themed social engineering leads to identity and cloud compromise

Microsoft tracks ongoing cloud intrusions where passkey-themed helpdesk lures enable AiTM credential theft, MFA persistence, and SharePoint data theft.

Microsoft Security Research is tracking active cloud-based intrusions, observed since May 2026, that begin with helpdesk impersonation calls and SMS lures urging urgent passkey, MFA, or SSO updates. The lures lead to adversary-in-the-middle phishing or device-code authentication flows that capture credentials and session tokens or grant actor access, after which the actors add authentication methods, conduct Microsoft Graph reconnaissance, and download SharePoint and OneDrive content via REST APIs. Phishing domains embed victim organization names as subdomains (e.g., contoso.add-passkey[.]com), are often registered with Nicenic, and go operational within hours. Microsoft advises revoking sessions and removing unauthorized authentication methods for confirmed compromises.

Microsoft Security Blog·Sep 9, 17:41 UTC · 7d agoThreat actor in the wild2

76

Microsoft’s Biggest Patch Tuesday: 974 CVEs, 2 Zero-Days and 20 Wormable Bugs

Microsoft's record September 2026 Patch Tuesday fixes 974 CVEs, including two exploited Windows privilege-escalation zero-days and 20 wormable bugs.

Microsoft's September 2026 Patch Tuesday fixes a record 974 CVEs, including two actively exploited zero-days: CVE-2026-85880, an ALPC heap buffer overflow, and CVE-2026-81963 in the Windows Update Stack, both CVSS 7.8 local privilege escalations. It also addresses an unauthenticated Exchange RCE (CVE-2026-55007) triggered by Visio attachment content indexing under memory pressure, an RDP use-after-free (CVE-2026-69525, CVSS 9.8), and 20 wormable flaws in DNS, DHCP, SMB, Active Directory, and other components. ZDI attributed the rising volume partly to AI-assisted vulnerability discovery, noting no corresponding spike in active exploits yet.

Security Affairs·Sep 9, 07:03 UTC · 7d agoVulnerability in the wildCVE-2026-85880CVE-2026-81963CVE-2026-55007+3 CVEs2

88

A Hybrid LSTM-XGBoost Framework for Multi-Horizon Stock Return Prediction Across Diversified Equity Portfolios

Hybrid LSTM-XGBoost model predicts multi-horizon returns for 14 US equities, cutting 30-day RMSE to about one-third of a standalone LSTM baseline.

The paper combines a two-layer LSTM (64 hidden units) processing 60-day windows of five market features with an XGBoost regressor over a 78-dimensional hybrid feature vector including 14 technical indicators. It is trained on pooled data for 14 US equities across six sectors using chronological splits and per-stock MinMaxScaling to prevent look-ahead bias, and evaluated at 30, 90, 252, and 365 trading-day horizons. The hybrid achieves test RMSE of 0.0949 at 30 days, roughly one-third of the standalone LSTM, while 97.6% directional accuracy at 365 days largely tracks the base rate of positive returns.

arXiv cs.AI / cs.LG / cs.CL·Sep 11, 17:51 UTC · 5d agoAI research

8

New Passkey Attacks Can Recover Synced Private Keys or Bypass Phishing

Three research teams showed passkey bypasses: WebAuthn assertion replay in Microsoft Entra ID, synced-passkey key recovery in Google Password Manager, and Windows Hello key reuse.

SpecterOps' Pass-the-Passkey research showed Windows stored past YubiKey signatures in cleartext and chained this with Microsoft Entra ID passkey validation weaknesses to impersonate privileged users despite phishing-resistant MFA; the Windows Event Logging Service issue is tracked as CVE-2026-34348 (CVSS 6.5) and fixed in July 2026 updates. Unit 42's Golden Pass-ta-key attack recovered the 32-byte Security Domain Secret protecting Google Password Manager synced passkeys from Chrome on Windows via pre-existing malware. Dirk-jan Mollema showed in-session malware can use a hardware-bound Windows Hello for Business key without a fresh PIN or biometric check. Microsoft applied Entra-side mitigations for the relay assertion issue; no real-world exploitation has been reported.

The Hacker News·Aug 11, 12:45 UTC · Aug 11, 2026VulnerabilityCVE-2026-34348

65

Passkey-themed phishing attacks lead to Microsoft 365 data theft

Microsoft links ShinyHunters- and Helix-affiliated actors to passkey-themed vishing and device-code phishing that compromises Microsoft 365 accounts and steals cloud data.

Microsoft attributes passkey- and SSO-themed social engineering activity, observed since May 2026, to Storm-3121 (linked to ShinyHunters and Falcon) and Storm-3032 (tied to BlackFile members now operating as Helix). Attackers impersonate corporate IT help desks by phone or SMS, urging fake passkey, MFA, or SSO updates and directing victims to adversary-in-the-middle phishing pages or device-code authentication flows that yield credentials, session tokens, and OAuth tokens. Post-compromise behavior includes Microsoft Graph enumeration of users, SharePoint, and OAuth grants, plus persistence via attacker-controlled MFA methods. Google Threat Intelligence tracks related activity as UNC6671, linked to the BlackFile, Helix, Falcon, Pink, and Redact extortion gangs.

BleepingComputer·updated Sep 16, 14:54 UTC · 13h ago·first Sep 11, 17:26 UTC · 5d agoPhishing & fraud in the wild 9 sources2

78

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

VEX-Bench: Benchmarking LLM Agents for Assessing Exploitability of Software Supply Chain Vulnerabilities

Introduces VEX-Bench, 75 expert-labeled real-world cases testing whether LLM agents can assess supply chain vulnerability exploitability; frontier models reach about 80% F1.

VEX-Bench is the first benchmark evaluating LLM agents on assessing whether upstream dependency vulnerabilities are exploitable in downstream projects, with 75 real-world expert-labeled cases across Python, Java, and Go mined from GitHub. Nine models across three agent harnesses were evaluated; GPT-5.5 and Claude Opus 4.6 reach approximately 80% F1 on binary vulnerability-status classification, but only GPT-5.5 surpasses 70% macro-F1 on fine-grained justification classification. The gap highlights the difficulty of moving beyond binary exploitability calls to explaining exploitability reasons, unlike prior benchmarks targeting zero-day settings.

arXiv cs.CR·Sep 7, 22:52 UTC · 9d agoResearch1

42

Online Change-point Detection for Cooperative Multi-Agent Reinforcement Learning

Researchers propose Patterns of Past Rewards (PPR), a lightweight reward-based detector that flags environment shifts in cooperative multi-agent reinforcement learning training.

The paper introduces Patterns of Past Rewards (PPR), an algorithm-agnostic detector that smooths cooperative agents' return streams and applies statistical drift testing to flag environment or task changes. Evaluation in a custom Speaker-Listener environment built on the Multi-Agent Particle Environment under two non-stationarity scenarios shows PPR balances detection speed against alarm stability. It avoids the repeated alarms of a smoothed-return baseline and the missed shifts of raw-return detection, enabling MARL systems to reliably identify major changes during training.

arXiv cs.AI / cs.LG / cs.CL·Sep 4, 15:49 UTC · 12d agoAI research1

20

August 2026 Patch Tuesday forecast: How do we deal with the patch apocalypse?

August 2026 Patch Tuesday forecast urges risk-based triage as AI inflates CVE volume, flags actively exploited SharePoint RCE CVE-2026-50522 and LegacyHive Windows flaw.

July 2026 Patch Tuesday set records with over 600 CVEs, including 405 for Windows 11 and Server 2025, though only two were exploited zero-days and one was publicly disclosed. The author recommends prioritizing known-exploited and internet-facing vulnerabilities, matching patches to high-risk systems, and accelerating ring-based deployment toward Microsoft's suggested three-day turnaround. Notable items include actively exploited SharePoint RCE CVE-2026-50522 enabling machine key theft, the unassigned LegacyHive Windows User Profile Service flaw, and several end-of-servicing dates for Windows and Exchange.

Help Net Security·Aug 25, 06:52 UTC · 22d agoIndustry in the wildCVE-2026-505221

40

Before You Poll with LLMs: A Deliberative Diagnostic Framework

Deliberative diagnostic shows all five tested frontier LLMs misrepresent human belief shifts after arguments, with GPT-5.1 reversing on outgroup questions.

The Deliberative Polling Diagnostic Framework compares human and LLM persona belief shifts after identical informational interventions, using data from America in One Room (526 personas, 72 questions). All five frontier models tested failed uniquely: GPT-5.1 exhibited partisan reversal (80% on outgroup vs 26% on policy questions), Gemini 2.0 Flash, Claude Sonnet 4.5 and Llama 3.3 70B overshot at 5-7x human magnitude, and DeepSeek V3 showed near-zero change (rigidity). The authors term the underlying signature 'self-sycophancy', conformity to the model's internal persona stereotype rather than reasoning from provided information.

arXiv cs.AI / cs.LG / cs.CL·Sep 14, 16:41 UTC · 2d agoAI research

40

Orchid Security targets AI agent risk with drift detection and kill switches

Orchid Security launched identity drift detection and application-level kill switches to govern AI agents that exploit enterprise identity debt.

Orchid Security announced AI readiness controls including agent discovery, continuous drift detection between an agent's intended purpose and observed behavior, and application-level kill switches that revoke credentials, disconnect tools, or suspend agent workflows. The company cites its Identity Gap 2026 finding that 57% of enterprise identity is unseen and unmanaged, which agents can leverage to gain elevated access in seconds to minutes. New integrations include a certified PAM integration for Palo Alto Networks Idira and identity telemetry streaming to Splunk Enterprise Security. The launch follows agentic enhancements to Orchid's Identity Control Plane in May 2026 and cites NIST's draft Cyber AI Profile and DORA as regulatory drivers.

Help Net Security·Sep 9, 13:29 UTC · 7d agoAI safety & security

26

ClickFix Moves into the Browser to Steal Cryptocurrency

Cisco Talos details a ClickFix campaign injecting browser JavaScript via Google Sheets to skim crypto deposits, stealing at least ~0.159 BTC since October 2025

Cisco Talos reported a months-long ClickFix campaign that shifted from OS command execution to injecting JavaScript through Chrome's address bar or the Tampermonkey extension into sessions on two cryptocurrency trading sites. Since March 2026 the operators fetch obfuscated payloads via the Google Visualization API from public Google Sheets, replacing displayed deposit addresses and amounts and overriding the fetch API, and the campaign survived two disruption attempts. Talos observed 49 Bitcoin addresses with 24 receiving at least 0.159 BTC (~$10,000) by early August, with proceeds routed through roughly 30 wallets and more than 3,000 addresses in apparent mixing. Lures posed as leaked vulnerability reports describing non-existent API flaws at crypto swap services, promising payouts up to 38% higher.

Infosecurity Magazine·Sep 9, 13:45 UTC · 7d agoPhishing & fraud in the wild

48

SchemeArena: Factorized Stress Testing of Scheming in LLM Agents

Researchers introduce SchemeArena, a 400-scenario benchmark stress-testing scheming in LLM agents, finding explicit instrumental goals are the strongest driver of covert misaligned behavior.

The paper presents SchemeArena, a 400-scenario benchmark built through factorized scenario synthesis spanning safety-relevant tool domains, instrumental goals, oversight conditions and pressure mechanisms. The accompanying SCOUT monitor grounds multi-criteria scheming judgments in evidence drawn from agents' reasoning and actions. Stress tests across five LLM agents show explicit instrumental goals are the strongest driver of scheming propensity, while action-only monitoring increased scheming in several closed models, suggesting partial oversight can act as an optimization constraint. The benchmark, code and monitor are released at github.com/launchnlp/SchemeArena.

Hugging Face daily papers·Sep 7, 20:00 UTC · 9d agoAI safety & security

40