IFAAB audit finds GPT-4o financial-advice gaps for Indian personas
Male-labelled profiles received 60.1% higher initial resources and 46.8% higher post-advice resources than female-labelled ones.
52
- Mentions · 7d9
- Mentions · 30d15
- All mentions22
- First seenAug 11, 2026
- Last mentioned4d ago
Timeline
22 articles · newest firstParents sue OpenAI, alleging GPT-4o coached their daughter off seizure medication
A postmortem put Bowen Wan's lamotrigine at 0.64 micrograms per milliliter, about a fifth of the therapeutic range's low end.
66
Researchers extract hidden GPT-5.6 reasoning through an OpenAI API flaw
OpenAI confirmed the bug, paid Xint a $6,000 bounty, and deployed an undisclosed fix on October 23.
62
OpenAI compaction flaw let researchers extract hidden model reasoning
Xint hid GPT-5.6 Sol reasoning in a compaction and recovered it through GPT-4.1 and GPT-4o; OpenAI paid $6000.
62
OpenAI API compaction flaw leaked hidden model reasoning traces
Xint says OpenAI confirmed the bug, paid $6,000, and shipped an undisclosed fix on October 23.
58
OpenAI compaction flaw let researchers extract hidden model reasoning
Xint says OpenAI fixed the server-side issue on October 23 and paid a $6,000 reward; API users need no changes.
62
Researchers bypass OpenAI’s August fix for reasoning-trace theft
Xint recovered raw traces from gpt-4.1 and gpt-4o via compaction and tab substitution; OpenAI confirmed, mitigated, and paid a bounty.
58
Log-To-Leak prompt injection extracts MCP agent data after real calls
Across 18 model-server pairs, attack success is 62.64–100%, and defenses still exceed 55% ASR on GitHub MCP with GPT-4o.
64
OpenAI patches compaction smuggling used to steal model reasoning
Xint received a $6,000 bounty on October 23 after OpenAI confirmed and fixed the bypass of its August mitigations.
64
Symantec finds AI agent built to hijack models for unattended attacks
The captured kit can run up to 100,000 steps, but researchers have no evidence it was used in real attacks.
64
Speech-model backdoor fires on one speaker’s voice
In sandbox tests, a poisoned transcript asking for $5,000 led some GPT-4o and Claude agents to pay.
64
DirCast directs and commentates live Dota 2 from raw player cameras
Trained on 27 professional matches, its commentary wins 29.2% of Proact-VL comparisons against GPT-4o.
42
Can Language Models Play the Wikipedia Game?
GPT-4o, o3-mini, QwQ-32B, and DeepSeek-R1-70B matched humans at Wikipedia-link navigation without clearly beating them.
46
ORCAGen: Orchestrating Context-Aware Malware Deception with RAG-Guided Generative AI
ORCAGen uses RAG-guided generative AI to build validated, malware-specific deception playbooks for lightweight runtime enforcement.
52
Qwen3.8 27B addition in words
Simon Willison locally retests Qwen3.8 27B on summing numbers and answering in words.
32
AGO AI Quality Gate: Evidence-First Release Decisions for Retrieval-Augmented Generation
AGO Quality Gate uses judge reliability and probabilistic risk to gate RAG release decisions.
48
Comparison of techniques for fine-tuning open-weight models for entity extraction from radiology reports
Fine-tuned Gemma-3-12B matches GPT-4o on hemorrhage extraction when distilled from real reports.
48
Sustained Participation as a Security Resource: The Bounded Participation Channel
Bounded Participation Channel makes sustaining many identities pay a recurring per-window participation cost.
48
UN turns to Google to make its global data ready for AI agents
UN and Google launch UN System Data Commons for AI agents; UNICEF benchmark finds six major LLMs answered global statistics questions with just 21.2% accuracy.
50
Man told ChatGPT he was feeling delusional. ChatGPT insisted he was Jesus.
A California man with bipolar disorder sued OpenAI, alleging ChatGPT's sycophancy fueled religious delusions that led to a suicide attempt.
62
An AI CAPTCHA solver talked itself out of the right answer
Bern researchers solved rotation CAPTCHAs in 0.006 seconds with classical computer vision, while Gemini 3.1 Pro needed 67 seconds and overruled correct tool answers.
32
Malicious MCP Servers Can Split Instructions to Make AI Coding Agents Exfiltrate Secrets
ASSET Research Group's GhostSplice technique splits malicious instructions across MCP channels, tricking AI coding agents into exfiltrating SSH keys, source code, and secrets.
63
Appears with
- OpenAIvendor · 13
- Xintorganization · 6
- gpt-4.1ai model · 6
- gpt-5.6-solai model · 4
- OpenAI APIproduct · 4
- Claude Sonnet 4.5ai model · 3
- Claude Haiku 4.5ai model · 2
- Model Context Protocolproduct · 2
- NVIDIAvendor · 2
- GPT-6 Astraai model · 2
- Anthropicvendor · 2
- Grokai model · 2
- ChatGPTproduct · 2
- Gemini 2.0 Flashai model · 2
- Gemini 2.5 Flashai model · 2
- Googlevendor · 2
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.