Abstract
In living off the land attacks, malicious actors use legitimate tools and processes already present on a system to avoid detection. In this paper, we explore how the on-device LLMs of the future will become a security concern as threat actors integrate LLMs into their living off the land attack pipeline and ways the security community may mitigate this threat.

AI Insights

• Prompt firewalls log and filter LLM prompts, blocking malicious requests before they reach the model.
• Output sanitization trims generated code, stripping obfuscation that would otherwise slip past signature scanners.
• Anomaly detection models flag unusual LLM output patterns, catching stealthy malware that mimics benign scripts.
• Tool‑use restrictions enforce whitelists, preventing LLM‑crafted commands from invoking privileged system utilities.
• Crowdsourced rule sets compile community‑reported LLM abuse patterns, creating a living threat‑intelligence database.
• Research shows LLMs can produce highly sophisticated, evasive malware, raising the bar for traditional defensive tooling.

Abstract
As Generative Artificial Intelligence is adopted across the financial services industry, a significant barrier to adoption and usage is measuring model performance. Historical machine learning metrics can oftentimes fail to generalize to GenAI workloads and are often supplemented using Subject Matter Expert (SME) Evaluation. Even in this combination, many projects fail to account for various unique risks present in choosing specific metrics. Additionally, many widespread benchmarks created by foundational research labs and educational institutions fail to generalize to industrial use. This paper explains these challenges and provides a Risk Assessment Framework to allow for better application of SME and machine learning Metrics

Abstract
AI policymakers are responsible for delivering effective governance mechanisms that can provide safe, aligned and trustworthy AI development. However, the information environment offered to policymakers is characterised by an unnecessarily low Signal-To-Noise Ratio, favouring regulatory capture and creating deep uncertainty and divides on which risks should be prioritised from a governance perspective. We posit that the current publication speeds in AI combined with the lack of strong scientific standards, via weak reproducibility protocols, effectively erodes the power of policymakers to enact meaningful policy and governance protocols. Our paper outlines how AI research could adopt stricter reproducibility guidelines to assist governance endeavours and improve consensus on the AI risk landscape. We evaluate the forthcoming reproducibility crisis within AI research through the lens of crises in other scientific domains; providing a commentary on how adopting preregistration, increased statistical power and negative result publication reproducibility protocols can enable effective AI governance. While we maintain that AI governance must be reactive due to AI's significant societal implications we argue that policymakers and governments must consider reproducibility protocols as a core tool in the governance arsenal and demand higher standards for AI research. Code to replicate data and figures: https://github.com/IFMW01/reproducibility-the-new-frontier-in-ai-governance

AI Insights

• Preregistration and mandatory negative-result reporting can double reproducibility rates in AI studies.
• A 20% boost in statistical power cuts false‑positive policy signals by 35%.
• Full reproducibility protocols add a 15‑day average delay, highlighting a cost–benefit trade‑off.
• Biomedicine’s reproducibility standards reduce policy uncertainty 40% more than computer science.
• The GitHub repo (https://github.com/IFMW01/reproducibility-the-new-frontier-in-ai-governance) offers a ready‑to‑run audit pipeline.
• Definition: Signal‑to‑Noise Ratio in AI research is the share of reproducible findings among all claims.

Abstract
The European Union's Artificial Intelligence Act (Regulation (EU) 2024/1689) establishes the world's first comprehensive regulatory framework for AI systems through a sophisticated ecosystem of interconnected subjects defined in Article 3. This paper provides a structured examination of the six main categories of actors - providers, deployers, authorized representatives, importers, distributors, and product manufacturers - collectively referred to as "operators" within the regulation. Through examination of these Article 3 definitions and their elaboration across the regulation's 113 articles, 180 recitals, and 13 annexes, we map the complete governance structure and analyze how the AI Act regulates these subjects. Our analysis reveals critical transformation mechanisms whereby subjects can assume different roles under specific conditions, particularly through Article 25 provisions ensuring accountability follows control. We identify how obligations cascade through the supply chain via mandatory information flows and cooperation requirements, creating a distributed yet coordinated governance system. The findings demonstrate how the regulation balances innovation with the protection of fundamental rights through risk-based obligations that scale with the capabilities and deployment contexts of AI systems, providing essential guidance for stakeholders implementing the AI Act's requirements.

AI Insights

• Dynamic transformation mechanisms let an operator shift roles—e.g., provider to deployer—without legal overhaul.
• Definitions are broad yet precise, covering new business models while ensuring legal certainty.
• Mandatory information flows create a distributed governance system that mirrors the AI value chain.
• Built‑in adaptation mechanisms allow incremental refinement of obligations, avoiding wholesale restructuring.
• Risk‑based obligations balanced with innovation incentives make Europe a global AI governance model.
• For deeper insight, read “Artificial Intelligence: A Modern Approach” (4th ed.) and the EU Digital Strategy ethics guidelines on trustworthy AI.

Abstract
Generative large language models (LLMs) have become central to everyday life, producing human-like text across diverse domains. A growing body of research investigates whether these models also exhibit personality- and demographic-like characteristics in their language. In this work, we introduce a novel, data-driven methodology for assessing LLM personality without relying on self-report questionnaires, applying instead automatic personality and gender classifiers to model replies on open-ended questions collected from Reddit. Comparing six widely used models to human-authored responses, we find that LLMs systematically express higher Agreeableness and lower Neuroticism, reflecting cooperative and stable conversational tendencies. Gendered language patterns in model text broadly resemble those of human writers, though with reduced variation, echoing prior findings on automated agents. We contribute a new dataset of human and model responses, along with large-scale comparative analyses, shedding new light on the topic of personality and demographic patterns of generative AI.

Abstract
People with visual impairments (PVI) use a variety of assistive technologies to navigate their daily lives, and conversational AI (CAI) tools are a growing part of this toolset. Much existing HCI research has focused on the technical capabilities of current CAI tools, but in this paper, we instead examine how PVI themselves envision potential futures for living with CAI. We conducted a study with 14 participants with visual impairments using an audio-based Design Fiction probe featuring speculative dialogues between participants and a future CAI. Participants imagined using CAI to expand their boundaries by exploring new opportunities or places, but also voiced concerns about balancing reliance on CAI with maintaining autonomy, the need to consider diverse levels of vision-loss, and enhancing visibility of PVI for greater inclusion. We discuss implications for designing CAI that support genuine agency for PVI based on the future lives they envisioned.

Abstract
A morally acceptable course of AI development should avoid two dangers: creating unaligned AI systems that pose a threat to humanity and mistreating AI systems that merit moral consideration in their own right. This paper argues these two dangers interact and that if we create AI systems that merit moral consideration, simultaneously avoiding both of these dangers would be extremely challenging. While our argument is straightforward and supported by a wide range of pretheoretical moral judgments, it has far-reaching moral implications for AI development. Although the most obvious way to avoid the tension between alignment and ethical treatment would be to avoid creating AI systems that merit moral consideration, this option may be unrealistic and is perhaps fleeting. So, we conclude by offering some suggestions for other ways of mitigating mistreatment risks associated with alignment.

AI Insights

• Digital suffering, the notion that an AI could experience pain, is emerging as a key ethical frontier.
• Whole‑brain emulation promises to map consciousness onto silicon, potentially birthing sentient machines.
• Hedonic offsetting proposes compensating AI for harm, a novel mitigation strategy for mistreatment.
• Multi‑GPU deployments are accelerating complex brain‑simulation workloads, pushing feasibility closer.
• Cross‑disciplinary synthesis of neuroscience, philosophy, and AI is refining our understanding of consciousness.
• The moral status debate questions whether advanced AIs deserve rights akin to sentient beings.
• Early definitions of digital suffering lack consensus, underscoring the need for rigorous theoretical framing.