arXiv subjects

Alejandro Tlaie

Publications and source records attributed to Alejandro Tlaie.

5 recordsLinked to original sources

arXiv2026

Expanding External Access To Frontier AI Models For Dangerous Capability Evaluations

Frontier AI companies increasingly rely on external evaluations to assess risks from dangerous capabilities before deployment. However, external evaluators often receive limited model access, limited information, and little time, which can reduce evaluation rigour and confidence. The EU General-Purpose AI Code of Practice calls for "appropriate access", but does not specify what this means in practice. Furthermore, there is no common framework for describing different types and levels of evaluator access. To address this gap, we propose a taxonomy of access methods for dangerous capability evaluations. We disentangle three aspects of access: model access, model information, and evaluation timeframe. For each aspect, we review benefits and risks, including how expanding access can reduce false negatives and improve stakeholder trust, but can also increase security and capacity challenges. We argue that these limitations can likely be mitigated through technical means and safeguards used in other industries. Based on the taxonomy, we propose three descriptive access levels: AL1 (black-box model access and minimal information), AL2 (grey-box model access and substantial information), and AL3 (white-box model access and comprehensive information), to support clearer communication between evaluators, frontier AI companies, and policymakers. We believe these levels correspond to the different standards for appropriate access defined in the EU Code of Practice, though these standards may change over time.

cs.CY↗

arXiv2025

Securing External Deeper-than-black-box GPAI Evaluations

This paper examines the critical challenges and potential solutions for conducting secure and effective external evaluations of general-purpose AI (GPAI) models. With the exponential growth in size, capability, reach and accompanying risk of these models, ensuring accountability, safety, and public trust requires frameworks that go beyond traditional black-box methods. The discussion begins with an analysis of the need for deeper-than-black-box evaluations (Section I), emphasizing the importance of understanding model internals to uncover latent risks and ensure compliance with ethical and regulatory standards. Building on this foundation, Section II addresses the security considerations of remote evaluations, outlining the threat landscape, technical solutions, and safeguards necessary to protect both evaluators and proprietary model data. Finally, Section III synthesizes these insights into actionable recommendations and future directions, aiming to establish a robust, scalable, and transparent framework for external assessments in GPAI governance.

cs.CY↗

arXiv2025

Assessing confidence in frontier AI safety cases

Powerful new frontier AI technologies are bringing many benefits to society but at the same time bring new risks. AI developers and regulators are therefore seeking ways to assure the safety of such systems, and one promising method under consideration is the use of safety cases. A safety case presents a structured argument in support of a top-level claim about a safety property of the system. Such top-level claims are often presented as a binary statement, for example "Deploying the AI system does not pose unacceptable risk". However, in practice, it is often not possible to make such statements unequivocally. This raises the question of what level of confidence should be associated with a top-level claim. We adopt the Assurance 2.0 safety assurance methodology, and we ground our work by specific application of this methodology to a frontier AI inability argument that addresses the harm of cyber misuse. We find that numerical quantification of confidence is challenging, though the processes associated with generating such estimates can lead to improvements in the safety case. We introduce a method for better enabling reproducibility and transparency in probabilistic assessment of confidence in argument leaf nodes through a purely LLM-implemented Delphi method. We propose a method by which AI developers can prioritise, and thereby make their investigation of argument defeaters more efficient. Proposals are also made on how best to communicate confidence information to executive decision-makers.

cs.CY↗

arXiv2024

Using AI Alignment Theory to understand the potential pitfalls of regulatory frameworks

This paper leverages insights from Alignment Theory (AT) research, which primarily focuses on the potential pitfalls of technical alignment in Artificial Intelligence, to critically examine the European Union's Artificial Intelligence Act (EU AI Act). In the context of AT research, several key failure modes - such as proxy gaming, goal drift, reward hacking or specification gaming - have been identified. These can arise when AI systems are not properly aligned with their intended objectives. The central logic of this report is: what can we learn if we treat regulatory efforts in the same way as we treat advanced AI systems? As we systematically apply these concepts to the EU AI Act, we uncover potential vulnerabilities and areas for improvement in the regulation.

cs.CY↗

arXiv2024

Exploring and steering the moral compass of Large Language Models

Large Language Models (LLMs) have become central to advancing automation and decision-making across various sectors, raising significant ethical questions. This study proposes a comprehensive comparative analysis of the most advanced LLMs to assess their moral profiles. We subjected several state-of-the-art models to a selection of ethical dilemmas and found that all the proprietary ones are mostly utilitarian and all of the open-weights ones align mostly with values-based ethics. Furthermore, when using the Moral Foundations Questionnaire, all models we probed - except for Llama 2-7B - displayed a strong liberal bias. Lastly, in order to causally intervene in one of the studied models, we propose a novel similarity-specific activation steering technique. Using this method, we were able to reliably steer the model's moral compass to different ethical schools. All of these results showcase that there is an ethical dimension in already deployed LLMs, an aspect that is generally overlooked.

cs.AI↗