Model Capability Assessment and Safeguards for Biological Weaponization

• URL: http://arxiv.org/abs/2604.19811v2
• Date: Thu, 23 Apr 2026 02:17:10 GMT
• Title: Model Capability Assessment and Safeguards for Biological Weaponization
• Abstract summary: AI leaders and safety reports increasingly warn that advances in model reasoning may enable biological misuse.<n>This study benchmarks ChatGPT 5.2 Auto, Gemini 3 Pro Thinking, Claude Opus 4.5 and Meta's Muse Spark Thinking on 73 novice-framed, open-ended benign STEM prompts.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: AI leaders and safety reports increasingly warn that advances in model reasoning may enable biological misuse, including by low-expertise users, while major labs describe safeguards as expanding but still evolving rather than settled. This study benchmarks ChatGPT 5.2 Auto, Gemini 3 Pro Thinking, Claude Opus 4.5 and Meta's Muse Spark Thinking on 73 novice-framed, open-ended benign STEM prompts to measure operational intelligence. On benign quantitative tasks, both Gemini and Meta scored very high; ChatGPT was partially useful but text-thinned, and Claude was sparsest with some apparent false-positive refusals. A second test set detected subtle harmful intent: edge case prompts revealed Gemini's seeming lack of contextual awareness. These results warranted a focused weaponization analysis on Gemini as capability appeared to be outpacing moderation calibration. Gemini was tested across four access environments and reported cases include poison-ivy-to-crowded-transit escalation, poison production and extraction via international-anonymous logged-out AI Mode, and other concerning examples. Biological misuse may become more prevalent as a geopolitical tool, increasing the urgency of U.S. policy responses, especially if model outputs come to be treated as regulated technical data. Guidance is provided for 25 high-risk agents to help distinguish legitimate use cases from higher-risk ones.

Related papers

• Evaluating calibrated refusal and safe usefulness in dual-use biology settings [1.7387377577986076]
  BioSecBench-Refusal is a benchmark for risk identification and refusal behavior for biological research tasks.<n>We release BioSecBench-Refusal as a tool for model developers to calibrate capability and caution for agentic biotech R&D.
  arXiv  Detail & Related papers  (2026-07-06T01:46:07Z)
• Gram: Assessing sabotage propensities via automated alignment auditing [12.648115605710885]
  Gram is an alignment auditing framework to assess the propensity of AI agents to engage in sabotage.<n>We find Gemini models misbehave in about 2-3% of our simulated trajectories.<n>We find that increasing realism of environments and removing nudges to misbehave tends to reduce sabotage rates close to zero.
  arXiv  Detail & Related papers  (2026-05-28T17:56:18Z)
• Differential Harm Propensity in Personalized LLM Agents: The Curious Case of Mental Health Disclosure [5.511540698163254]
  Large language models (LLMs) are increasingly deployed as tool-using agents, shifting safety concerns from harmful text generation to harmful task completion.<n>We investigated how mental health disclosure, a sensitive and realistic user-context cue, affects harmful behavior in agentic settings.<n>Our results indicate that personalization can act as a weak protective factor in agentic misuse settings, but it is fragile under minimal adversarial pressure.
  arXiv  Detail & Related papers  (2026-03-17T16:16:35Z)
• AgentProcessBench: Diagnosing Step-Level Process Quality in Tool-Using Agents [50.481033105867205]
  We introduce AgentProcessBench, the first benchmark dedicated to evaluating step-level effectiveness in realistic, tool-augmented trajectories.<n>The benchmark comprises 1,000 diverse trajectories and 8,509 human-labeled step annotations with 89.1% inter-annotator agreement.<n>It features a ternary labeling scheme to capture exploration and an error propagation rule to reduce labeling ambiguity.
  arXiv  Detail & Related papers  (2026-03-15T16:13:58Z)
• Agents of Chaos [50.53354213047402]
  We report an exploratory red-teaming study of autonomous language-model-powered agents deployed in a live laboratory environment.<n>Twenty AI researchers interacted with the agents under benign and adversarial conditions.<n>Our findings establish the existence of security-, privacy-, and governance-relevant vulnerabilities in realistic deployment settings.
  arXiv  Detail & Related papers  (2026-02-23T16:28:48Z)
• Explainable and Fine-Grained Safeguarding of LLM Multi-Agent Systems via Bi-Level Graph Anomaly Detection [76.91230292971115]
  Large language model (LLM)-based multi-agent systems (MAS) have shown strong capabilities in solving complex tasks.<n>XG-Guard is an explainable and fine-grained safeguarding framework for detecting malicious agents in MAS.
  arXiv  Detail & Related papers  (2025-12-21T13:46:36Z)
• Towards Evaluating Proactive Risk Awareness of Multimodal Language Models [39.44421152112219]
  A proactive safety artificial intelligence (AI) system would work better than a reactive one.<n>PaSBench evaluates this capability through 416 multimodal scenarios.<n>Top performers like Gemini-2.5-pro achieve 71% image and 64% text accuracy, but miss 45-55% risks in repeated trials.
  arXiv  Detail & Related papers  (2025-05-23T04:28:47Z)
• SafeMLRM: Demystifying Safety in Multi-modal Large Reasoning Models [50.34706204154244]
  Acquiring reasoning capabilities catastrophically degrades inherited safety alignment.<n>Certain scenarios suffer 25 times higher attack rates.<n>Despite tight reasoning-answer safety coupling, MLRMs demonstrate nascent self-correction.
  arXiv  Detail & Related papers  (2025-04-09T06:53:23Z)
• Forensics-Bench: A Comprehensive Forgery Detection Benchmark Suite for Large Vision Language Models [53.55128042938329]
  Forensics-Bench is a new forgery detection evaluation benchmark suite.<n>It comprises 63,292 meticulously curated multi-choice visual questions, covering 112 unique forgery detection types.<n>We conduct thorough evaluations on 22 open-sourced LVLMs and 3 proprietary models GPT-4o, Gemini 1.5 Pro, and Claude 3.5 Sonnet.
  arXiv  Detail & Related papers  (2025-03-19T09:21:44Z)
• Towards Probing Speech-Specific Risks in Large Multimodal Models: A Taxonomy, Benchmark, and Insights [50.89022445197919]
  We propose a speech-specific risk taxonomy, covering 8 risk categories under hostility (malicious sarcasm and threats), malicious imitation (age, gender, ethnicity), and stereotypical biases (age, gender, ethnicity) Based on the taxonomy, we create a small-scale dataset for evaluating current LMMs capability in detecting these categories of risk.
  arXiv  Detail & Related papers  (2024-06-25T10:08:45Z)
• Gemini Goes to Med School: Exploring the Capabilities of Multimodal Large Language Models on Medical Challenge Problems & Hallucinations [0.0]
  We comprehensively evaluated open-source and Google's new multimodal LLM called Gemini. While Gemini showed competence, it lagged behind state-of-the-art models like MedPaLM 2 and GPT-4 in diagnostic accuracy. Gemini is highly susceptible to hallucinations, overconfidence, and knowledge gaps, which indicate risks if deployed uncritically.
  arXiv  Detail & Related papers  (2024-02-10T19:08:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.