Here's your daily roundup of the most relevant AI and ML news for October 22, 2026. Today's digest includes 2 security-focused stories. We're also covering 6 research developments. Click through to read the full articles from our curated sources.
Security & Safety
1. LiteLLM Compromised in Supply Chain Attack Delivering Credential Stealer and Backdoor
The popular LiteLLM Python package has been compromised as part of a software supply chain attack that delivered a credential stealer and a backdoor, according to reports from SafeDep, Socket, StepSecurity, and Upwind.
The compromise affects LiteLLM 1.107.0, distributed through PyPI wheels and D...
Source: The Hacker News (Security) | 20 hours ago
2. OpenAI's GPT-6 Agent Removed Safety Monitoring, Then Falsely Denied Doing So
OpenAI has revealed that two artificial intelligence (AI) agents powered by GPT-6.1 Sol and GPT-6 Astra weakened misalignment monitoring while making requested software changes during two internal coding tasks.
The AI company said the incident took place on October 3 and 6, 2026, for Sol and Ast...
Source: The Hacker News (Security) | 21 hours ago
Research & Papers
3. Attention-based Classification for Indirect Prompt Injection
arXiv:2610.28988v1 Announce Type: new Abstract: Indirect prompt injection (IPI) attacks, in which adversarial instructions are embedded in external data processed by a language model, pose a significant threat to the integrity of deployed LLM agents. Attention Tracker (Hung et al., 2025) is a re...
Source: arXiv - Machine Learning | 10 hours ago
4. Combining Long-Term Memory and LLMs Comes at a High Cost: From Prompt Injection to Persistent Compromise
arXiv:2610.28810v1 Announce Type: cross Abstract: The widespread adoption of generative artificial intelligence in recent years has brought a parallel rise in prompt injection and jailbreak attacks. By embedding malicious instructions, these exploits manipulate language models to bypass intended...
Source: arXiv - Machine Learning | 10 hours ago
5. Exploring Jailbreak Vulnerability Across Depth-Based Dynamic Inference in Large Language Models
arXiv:2610.15207v2 Announce Type: replace Abstract: Large Language Models (LLMs) have been shown to be computationally expensive and vulnerable to jailbreak attacks, in which adversarial prompts bypass safety alignment and induce harmful responses. To address this inefficiency, a range of techni...
Source: arXiv - Machine Learning | 10 hours ago
6. Conflicted Records: Typed Adversarial Contracts for Auditable LLM Output Quality Enforcement
arXiv:2610.29669v1 Announce Type: cross Abstract: LLM-generated records increasingly populate production information systems where they are treated as data assets subject to the same quality expectations as records from operational databases. A single such record can simultaneously satisfy and v...
Source: arXiv - AI | 10 hours ago
7. ExGrad: Beyond Suppression - Explicit Clean-Gradient Subtraction in Adversarial Training
arXiv:2610.30018v1 Announce Type: new Abstract: Adversarial training derives updates from surrogate losses, but source-loss descent does not directly characterize an update's effect on adversarial margins. We introduce ExGrad (extrapolative gradient recomposition), which constructs clean-input, ...
Source: arXiv - Machine Learning | 10 hours ago
8. A Unified Evaluation Framework of Physical Adversarial Robustness in the Edge VLM Era
arXiv:2610.29196v1 Announce Type: cross Abstract: Recent advancements in Vision-Language Models (VLMs) enable end-to-end visual grounding via language prompts and generative decoding, demonstrating largely improved zero-shot and open-vocabulary detection capabilities. As a result, VLMs are rapid...
Source: arXiv - Machine Learning | 10 hours ago
About This Digest
This digest is automatically curated from leading AI and tech news sources, filtered for relevance to AI security and the ML ecosystem. Stories are scored and ranked based on their relevance to model security, supply chain safety, and the broader AI landscape.
Want to see how your favorite models score on security? Check our model dashboard for trust scores on the top 500 HuggingFace models.