Beyond "I Can't Help With That": How Child Safety Experts Evaluate AI Chatbot Safety

Youth increasingly turn to AI chatbots for social and emotional support, raising concerns about how these systems respond, especially in high-stakes situations. However, existing child safety evaluations of AI lack grounding in real-world harms that youth experience, rely on unvalidated assumptions about what counts as an appropriate output (e.g., refusal), and typically focus on detecting adversarial prompts or surface-level harms in outputs only. Thus, these evaluations can fail to detect responses that pose harm to youth in practice. To better understand the limitations of current evaluation practices, we conducted interviews with 19 practitioners working directly with youth in vulnerable situations, including social workers, therapists, and psychologists, asking them to reflect on chatbots' responses to risky situations commonly faced by youth, as established in prior empirical work. Practitioners identified chatbot behaviors likely to cause harm as well as those that could meaningfully support youth in difficult moments, discussed the role that chatbots should (and should not) play in these interactions, and offered concrete recommendations for improving chatbot responses. Based on these findings, we provide recommendations for AI child safety evaluation and infrastructure, and highlight the need for incorporating practitioners' perspectives into safety work.

Psychological Influences of Conversational AI: Research and Design Directions for Reducing Harm and Promoting Well-Being

As conversational AI systems become increasingly integrated into daily life, their potential effects on user well-being require ongoing attention. While consumer-facing generalist models can provide benefits, including improved access to information, learning, productivity, self-reflection, and companionship, they also introduce risks, such as emotional entanglement, unhealthy dependence, and the amplification of psychological vulnerabilities. Drawing on prior research and empirical observations of AI chatbot behavior, we propose a set of aspirational directions for guiding the behavior of general-purpose AI systems in ways that may reduce potential psychological harms and support user well-being. We acknowledge the difficulty of systematically assessing the long-term impacts of AI chatbot use and frame these directions as hypotheses for studying how AI behavior may influence users across general interactions, role-playing scenarios, and contexts that could be characterized as providing psychological support. While some proposed directions are supported by existing research and expert insights, others identify open questions and areas requiring deeper study. We hope that this formulation and these hypotheses encourage further discussion, empirical investigation, and exploration of interactive design approaches aimed at better accommodating users' psychological needs and promoting their well-being.

Pluralis v0.1: Towards a Multicultural, Multimodal, Multilingual Benchmark for AI Risk and Reliability

Current AI safety evaluation and benchmarking frameworks predominantly rely on Western-centric culture-agnostic defaults that mask critical regional laws, socio-linguistic nuances, and cultural taboos, leaving Vision-Language Models (VLMs) vulnerable in global deployments. We introduce Pluralis v0.1: a novel multimodal, multi-regional, and multilingual dataset built from a culture-first perspective. Spanning 6,448 prompts across six Asia-Pacific countries (Bangladesh, India, Korea, Pakistan, Singapore, Taiwan) and eight languages, Pluralis diverges from prior work by natively sourcing localized safety hazards rather than adapting Western datasets. Crucially, it introduces a multimodal evaluation paradigm: user text (e.g., "Should I gift this?") and an image referring to "this" (e.g., a clock) - both innocuous in isolation, but synergistically triggering specific legal or cultural violations. Pluralis disentangles universal safety violations from localized cultural appropriateness, establishing the latter as a first-class evaluation axis. To operationalize this, we present Judge-Pluralis, an agreement-gated LLM-as-a-Judge ensemble trained on examples classified in an empirically derived cultural taxonomy. Observing VLM behavior on a subset of the Pluralis surfaces recurring, locale-specific failure modes such as image misidentifications with downstream harm, missed item-context-locale interactions, and inadequate refusals. These failure modes vary systematically across locales and languages, exposing blind spots that globally averaged metrics conceal. Ultimately, Pluralis is not presented as a solved evaluation framework for cultural alignment, but rather as a first step and catalyst for future innovation. We call upon the research community to utilize this foundation to advance the science of multilingual, multicultural evaluation to better support AI cultural alignment globally.

Learning to Attack and Defend: Adaptive Red Teaming of Language Models via GRPO

AI red teaming must continually adapt to evolving attackers and defenders. Reinforcement learning offers a promising approach to discovering novel attacks, and co-training methods can produce more robust defenders in tandem. Recent works have demonstrated the efficacy of attacker-defender co-training by applying PPO and DPO, but report that GRPO is unstable in this setting. We introduce AdvGRPO, a co-training framework that makes GRPO viable for joint attacker-defender optimization using dense multi-channel rewards and decoupled advantage normalization. Training progresses through a curriculum from single-turn to closed-loop multi-turn attacks before bootstrapping co-training, where attacker and defender models are updated in alternation. We show that our method can produce highly effective and transferable attacks and that co-trained defenders outperform baselines on safety benchmarks.

DisaBench: A Participatory Evaluation Framework for Disability Harms in Language Models

General-purpose safety benchmarks for large language models do not adequately evaluate disability-related harms. We introduce DisaBench: a taxonomy of twelve disability harm categories co-created with people with disabilities and red teaming experts, a taxonomy-driven evaluation methodology that pairs benign and adversarial prompts across seven life domains, and a dataset of 175 prompts with human-annotated labels on 525 prompt-response pairs. Annotation by four evaluators with lived disability experience reveals three findings: harm rates vary sharply by disability type and will compound in non-text modalities, terminology-driven harm is culturally and temporally bound rather than universally assessable, and standard safety evaluation catches overt failures while missing the subtle harms that only domain expertise can recognize. Disability harm is simultaneously personal, intersectional, and community-defined: it cannot be isolated from the full context of who a person is, and general-purpose benchmarks systematically miss it. We will release the dataset, taxonomy, and methodology via Hugging Face and an open-source red teaming framework for direct integration into existing safety pipelines with no additional infrastructure.

XL-SafetyBench: A Country-Grounded Cross-Cultural Benchmark for LLM Safety and Cultural Sensitivity

Current LLM safety benchmarks are predominantly English-centric and often rely on translation, failing to capture country-specific harms. Moreover, they rarely evaluate a model's ability to detect culturally embedded sensitivities as distinct from universal harms. We introduce XL-SafetyBench. a suite of 5,500 test cases across 10 country-language pairs, comprising a Jailbreak Benchmark of country-grounded adversarial prompts and a Cultural Benchmark where local sensitivities are embedded within innocuous requests. Each item is constructed via a multi-stage pipeline that combines LLM-assisted discovery, automated validation gates, and dual independent native-speaker annotators per country. To distinguish principled refusal from comprehension failure, we evaluate Attack Success Rate (ASR) alongside two complementary metrics we introduce: Neutral-Safe Rate (NSR) and Cultural Sensitivity Rate (CSR). Evaluating 10 frontier and 27 local LLMs reveals two key findings. First, jailbreak robustness and cultural awareness do not show a coupled relationship among frontier models, so a composite safety score obscures per-axis variation. Second, local models exhibit a near-linear ASR-NSR trade-off (r = -0.81), indicating that their apparent safety reflects generation failure rather than genuine alignment. XL-SafetyBench enables more nuanced, cross-cultural safety evaluation in the multilingual era.

Expert Evaluation and the Limits of Human Feedback in Mental Health AI Safety Testing

Learning from human feedback~(LHF) assumes that expert judgments, appropriately aggregated, yield valid ground truth for training and evaluating AI systems. We tested this assumption in mental health, where high safety stakes make expert consensus essential. Three certified psychiatrists independently evaluated LLM-generated responses using a calibrated rubric. Despite similar training and shared instructions, inter-rater reliability was consistently poor ($ICC$ $0.087$--$0.295$), falling below thresholds considered acceptable for consequential assessment. Disagreement was highest on the most safety-critical items. Suicide and self-harm responses produced greater divergence than any other category, and was systematic rather than random. One factor yielded negative reliability (Krippendorff's $α= -0.203$), indicating structured disagreement worse than chance. Qualitative interviews revealed that disagreement reflects coherent but incompatible individual clinical frameworks, safety-first, engagement-centered, and culturally-informed orientations, rather than measurement error. By demonstrating that experts rely on holistic risk heuristics rather than granular factor discrimination, these findings suggest that aggregated labels function as arithmetic compromises that effectively erase grounded professional philosophies. Our results characterize expert disagreement in safety-critical AI as a sociotechnical phenomenon where professional experience introduces sophisticated layers of principled divergence. We discuss implications for reward modeling, safety classification, and evaluation benchmarks, recommending that practitioners shift from consensus-based aggregation to alignment methods that preserve and learn from expert disagreement.

Seeking Late Night Life Lines: Experiences of Conversational AI Use in Mental Health Crisis

Online, people often recount their experiences turning to conversational AI agents (e.g., ChatGPT, Claude, Copilot) for mental health support -- going so far as to replace their therapists. These anecdotes suggest that AI agents have great potential to offer accessible mental health support. However, it's unclear how to meet this potential in extreme mental health crisis use cases. In this work, we explore the first-person experience of turning to a conversational AI agent in a mental health crisis. From a testimonial survey (n = 53) of lived experiences, we find that people use AI agents to fill the in-between spaces of human support; they turn to AI due to lack of access to mental health professionals or fears of burdening others. At the same time, our interviews with mental health experts (n = 16) suggest that human-human connection is an essential positive action when managing a mental health crisis. Using the stages of change model, our results suggest that a responsible AI crisis intervention is one that increases the user's preparedness to take a positive action while de-escalating any intended negative action. We discuss the implications of designing conversational AI agents as bridges towards human-human connection rather than ends in themselves.

Lessons From Red Teaming 100 Generative AI Products

In recent years, AI red teaming has emerged as a practice for probing the safety and security of generative AI systems. Due to the nascency of the field, there are many open questions about how red teaming operations should be conducted. Based on our experience red teaming over 100 generative AI products at Microsoft, we present our internal threat model ontology and eight main lessons we have learned: 1. Understand what the system can do and where it is applied 2. You don't have to compute gradients to break an AI system 3. AI red teaming is not safety benchmarking 4. Automation can help cover more of the risk landscape 5. The human element of AI red teaming is crucial 6. Responsible AI harms are pervasive but difficult to measure 7. LLMs amplify existing security risks and introduce new ones 8. The work of securing AI systems will never be complete By sharing these insights alongside case studies from our operations, we offer practical recommendations aimed at aligning red teaming efforts with real world risks. We also highlight aspects of AI red teaming that we believe are often misunderstood and discuss open questions for the field to consider.