Results for “model”

Simon Willison SI

llm 0.37

Release: llm 0.37 There's a lot of stuff in this new release of my LLM CLI tool and Python library, representing around two months of accumulated features and fixes. I added gpt-6.1-sol and gpt-6.1-terra, plus a new llm openai endpoint add command to make it easier to configure additional models that use OpenAI-compatible endpoints. I switched the default model (if you have not set your own default) to gpt-6-luna - half the price of the previous default, gpt-5.6-luna. LLM's…

Read at Simon Willison

GeekWire Kurt Schlosser SI

Accent bias: AI models penalize certain speakers in interviews and tests, UW study finds

Models from Google and OpenAI scored American and British speakers higher than Chinese and Nigerian speakers, even when everyone read the same script.

Read at GeekWire

SiliconANGLE AI Duncan Riley SI

42AI launches with $62.5M to predict how traders act when markets turn volatile

Israeli artificial intelligence startup 42AI launched today with $62.5 million in funding to build what it calls a behavioral world model. Its first customers are brokerages that want to know how individual traders will act when markets turn volatile. 42AI offers software that estimates the odds of each choice a person might make in a

Read at SiliconANGLE AI

Buenos Aires Times Ignacio Olivera Doll, Bloomberg Geopolitics

Milei's boom-and-bust economy sparks exodus from Buenos Aires

La Libertad Avanza leader's economic model is creating new winners and losers – pushing his political popularity to the limit ahead of his re-election bid next year. Leer más

Read at Buenos Aires Times

Ars Technica Kyle Orland SI

Claude’s new rules against game cloning won’t stop the AI copying boom

The tsunami of imitators is way bigger than any one model maker can stem.

Read at Ars Technica

TechCrunch Startups Ram Iyer Startups

42AI comes out of stealth with $62.5M and a world model that can ‘predict human behavior’

Here's a question straight out of sci-fi: What if AI could predict what you were going to do before you decided to, even in a completely new situation?

Read at TechCrunch Startups

The Decoder Maximilian Schreiner SI

One brutal Minecraft night scared 50 AI agents into inventing chicken handoff protocols, a Fertilizer Officer, and fishing shifts

In experiments by Y Combinator Labs and Olam Labs, OpenAI's GPT-6.1 Sol model showed two very different sides. When 50 agents were tasked with peaceful civilization-building in Minecraft, they froze up in absurd bureaucracy. In direct competition against Claude Opus 5.5, they independently downloaded the game's source code to win. Simulations like these are meant to serve as an early warning system before AI swarms act unpredictably in real-world systems. The article One…

Read at The Decoder

9to5Google Ben Schoon

Gemini 4 Carbon lags just behind Claude in benchmark, but wins in value

Google’s new frontier model, Gemini 4 Carbon, brings some welcome improvements and while it doesn’t top the charts in a new benchmark, it does provide much more bang for your buck.

Read at 9to5Google

SiliconANGLE AI Ryan Stevens SI

Enterprises bring AI to their data as privacy rules tighten and costs climb

Every enterprise AI strategy eventually comes down to the same question: Should proprietary data travel to the model, or should the model come to the data? As privacy rules tighten and geopolitical risk rises, more organizations are choosing the second option. That is changing how data platform providers present themselves, and Cloudera Inc. has built

Read at SiliconANGLE AI

TestingCatalog Dwaybe Cubbins SI

GPT-6.2 Astra appears in Codex usage ahead of possible launch

GPT-6.2-astra is showing up in Codex usage analytics for several users, hinting OpenAI is quietly serving a new Astra model.

Read at TestingCatalog

Rest of World Rina Chandran and Viola Zhou SI

Startups recalibrate Chinese AI for the West

A growing number of startups are editing powerful Chinese open-weight models for Western businesses that are concerned about censorship and security.

Read at Rest of World

arXiv (research papers) Xu Zhang SI

ATrust: Adaptive Trust Boundary Expansion for Securing LLM Agents

As large language model (LLM) agents take on more complex tasks with greater autonomy, security has become a major obstacle to their broader deployment. Unlike in attacks that require complex exploitation of operating system vulnerabilities, attackers can easily hijack agents through indirect prompt injection (IPI), leading to privacy leakage, financial loss, and destructive operations. However, existing defenses often make security decisions based on limited static trusted…

Read at arXiv (research papers)

arXiv (research papers) Kainan Zhou SI

SearchInject: Retrieval Exposure Control for Indirect Prompt Injection in Language-Model Search

An indirect prompt injection can affect a search-augmented large language model only after its carrier page reaches model context. SearchInject controls this admission decision with a joint score combining retrieval relevance, search-abuse cues, instruction likelihood, and two relevance interactions. A quarantine threshold is selected under an empirical 0.5% clean-candidate false-quarantine constraint on a calibration split. Evaluation uses 300 held-out SciFact queries and…

Read at arXiv (research papers)

Tech in Asia Adinda Pryanka Startups

Replacing expensive A/B tests with AI models

Pre-launch ad testing can cost thousands. Voygic uses AI predictive models to give consumer brands highly accurate feedback in minutes.

Read at Tech in Asia

arXiv (research papers) Christopher Krakauer SI

Where Does MCP Security End? A Three-Tier Mitigation Mapping Framework for the OWASP MCP Top 10

The Model Context Protocol (MCP) has become a leading standard for connecting Large Language Models (LLMs) to external tools, software, and data sources. MCP architecture favors ease of integration over security, blending the data and control planes into a single natural language channel. This introduces confused deputy and prompt injection vulnerabilities that cannot be completely mitigated in the software development lifecycle (SDLC) of an MCP server. Structured security…

Read at arXiv (research papers)

arXiv (research papers) Haowen Dai SI

StateFence: Task-State Protection Against Evidence-Chain Contamination in Blue-Team LLM Agents

Indirect prompt injection creates a distinctive risk for blue-team Large Language Model (LLM) agents in Security Operations Centers (SOCs). Malicious instructions can be embedded in telemetry, tickets, retrieval results, or memory and then propagate through alert triage, incident summarization, response recommendation, and Indicator of Compromise (IOC) extraction as locally plausible intermediate decisions. We characterize this failure mode as evidence-chain contamination of…

Read at arXiv (research papers)

arXiv (research papers) Junwei Yao SI

Ranked but Not Stopped: Detecting Prompt-Injection Hijacks from Typed Decision Models' Output Probabilities

Programs that use a language model as a classifier act on its most probable label. Text injected into the input can move that label to one the attacker chooses, a hijack. A monitor that does not know the attacker's label can read the model's own confidence and raise an alarm when the top probability is low. We test this monitor on AG News topic classification with five systems, three of them served through a provider's API. Pooled over five families of injected text, low…

Read at arXiv (research papers)

Tech in Asia Glenn Kaonang Startups

Big AI cashes in on its own risks

Anthropic, OpenAI, and Google are selling defenses against risks their models helped create. For clients, it means even higher cybersecurity budgets.

Read at Tech in Asia

arXiv (research papers) Sreehari Sankar SI

LEANGUARD: Lean-Verified Guardrails for Tool-Using Agents

Tool-using language-model agents often execute actions that their domain policies forbid. We present LeanGuard, a runtime that checks every proposed tool call against a policy written in Lean4 before the call reaches the environment. Policies combine state predicates with scoped past-time temporal operators over the execution history, evaluation is fail-closed, and the decision procedure is proved equivalent to a LeanLTL semantics. A denied call can be corrected in two ways…

Read at arXiv (research papers)

Journal articles (OpenAlex) Christopher Salame, Mohammad Antor, Fatiha Djebbar et al. · Computers

Cross-Tier ML Intrusion Detection for Resource-Constrained Environmental IoT Sensor Networks

Compromised Internet of Things (IoT) sensor networks in critical infrastructure such as water-quality monitoring carry direct public-health consequences, yet intrusion-detection research rarely combines threat modelling, hardware-constrained deployment and regulatory mapping. This study applies the STRIDE framework to the 14 attack classes of the Edge-IIoTset dataset, evaluates six classifiers across a Raspberry Pi 4 gateway tier and an ESP32-C6 sensor tier, and maps…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) José María González de Durana, Luis Carlos Rabelo, Minas Pantelidakis et al. · Energies

A Conceptual Framework for Cybersecurity Modeling in IoT-Enabled Microgrids Based on System Dynamics, Agent-Based Modeling, and Statecharts

The integration of Internet of Things (IoT) devices in microgrids enhances efficiency and adaptability but also introduces significant cybersecurity risks. This paper presents a conceptual hybrid modeling framework for cybersecurity modeling in IoT-enabled microgrids by combining System Dynamics (SD), Agent-Based Modeling (ABM), and statecharts. SD captures energy flows and storage dynamics, ABM models decentralized interactions among microgrid components, and statecharts…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Xiaokai Zhou, Hao Huang, Quanqing Xu et al. · Data Science and Engineering Tech policy

VFCI: A Federated Framework for Vertical Causal Inference

Causal inference is crucial for understanding the relationship between cause and effect and is widely applied in fields such as medicine and policy decision-making. When the input data is distributed over multiple participants, most causal inference models require data to be centrally aggregated. However, participants are often prohibited or reluctant to share sensitive data due to data protection regulations and privacy concerns. To address this, federated learning has been…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Jawaher Abdulwahab Fadhil, Marwan Aziz Mohammed · Egyptian Informatics Journal Tech policy

Optimized hybrid cryptographic model with deep learning for secure IoT data

The rapid growth of smart devices has increased the demand for robust and scalable encryption mechanisms while expanding the attack surface to cyber threats, especially in Internet of Things (IoT) environments. Recent advances in Deep Learning (DL) have created new opportunities to improve encryption methods and enable secure data storage and transmission against emerging cyber threats. This paper presents an optimized security framework that integrates DL with an optimized…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Zakeyeh Aftabi, Mohammadnabi Jalali, Hossein Salehi et al. · Results in Engineering Geopolitics

Coupling hydroclimatic evidence and water conflict analysis: A TWINS–GMCR+ assessment of Iran–Iraq hydropolitical relations in the Tigris–Euphrates Transboundary Hydrosystem

The Tigris–Euphrates Transboundary Hydrosystem (TETH) is shaped by hydroclimatic drivers, large-scale water infrastructure, and complex hydropolitical relations among riparian states. This study uses a multi-stage analytical framework that combines independent hydroclimatic and hydrological evidence with the Transboundary Water Interaction Nexus (TWINS), the Graph Model for Conflict Resolution (GMCR+), and post-equilibrium social-choice screening to examine Iran–Iraq…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Qiulin Wu, Guanghai Li · PLoS ONE Geopolitics

A quantitative assessment of Chinese county high school policies using the PMC index model

The county high school policies (CHSPs) is crucial for improving the quality of basic education in China's counties and promoting rural social development. As policy documents are the foundational elements of policy systems, it is crucial to evaluate current policy texts, identify their shortcomings, and provide empirical evidence for future policy development. This study collected 55 policy texts from provincial and municipal governments, utilizing the Policy Modeling…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Yassine El Hajoui, Nabil El BAOUCHARI · Discover Global Society Geopolitics

Global economic consequences of the fifth global trade war under Trump 2.0 evidence from a dynamic multiregional computable general equilibrium model

The objective of this article is to analyze the impact of the strategic trade war initiated by the United States on April 2, 2025, dubbed “Liberation Day,” and to assess its respective effects on the economies involved. This protectionist measure, aimed at rebalancing the U.S. trade balance, was quickly suspended for all countries except China for a period of 90 days, turning tariffs into a bilateral negotiation lever. To study these dynamics, we use a dynamic multi-regional…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Xuan Wu, Guangjian Li, Zhixin Zhang · Journal of Comparative Policy Analysis Research and Practice Geopolitics

A Multidimensional Comparison of Artificial Intelligence Policy Tools in China, the United States, and the European Union: An Analysis Based on Large Language Models

This study introduces an LLMs-enabled policy computing method to quantitatively compare AI policy instruments across China, the US, and the EU based on a supply–environment–demand taxonomy. It identifies three distinct paradigms. China follows a supply-driven model focused on state-led capacity building. The US adopts a balanced strategy linking technological leadership with regulatory influence. The EU pursues a regulation-centric pathway prioritizing legal frameworks. Each…

Read at Journal articles (OpenAlex)

Older stories