LIVE PULSE
3.9 Anthropic CEO Amodei calls for slower AI development and shared safety rules — 11 src2.1 Agility Robotics unveils Digit 5 humanoid for warehouses and factories — 2 src2.0 Apple ships rebuilt Siri with Google Gemini, but not in the EU — 2 src1.7 Siri AI in macOS 27 Golden Gate: FAQ, Germany availability, privacy questions — 2 src1.4 Sam Altman says OpenAI will not go public in 2026, citing AI safety concerns — 5 src1.1 OpenAI contractors review real ChatGPT conversations to rate responses, report says — 2 src1.1 Anthropic data retention policy prompts firms to limit Claude use for sensitive work — 1 src1.1 Study examines issue bias in LLMs used as writing assistants before Swedish 2026 election — 1 src1.1 Study Audits Misalignment in Multi-Modal World Models — 1 src1.1 Retrieval-Grounded Reasoning Approach Proposed for Universal Multimodal Embeddings — 1 src3.9 Anthropic CEO Amodei calls for slower AI development and shared safety rules — 11 src2.1 Agility Robotics unveils Digit 5 humanoid for warehouses and factories — 2 src2.0 Apple ships rebuilt Siri with Google Gemini, but not in the EU — 2 src1.7 Siri AI in macOS 27 Golden Gate: FAQ, Germany availability, privacy questions — 2 src1.4 Sam Altman says OpenAI will not go public in 2026, citing AI safety concerns — 5 src1.1 OpenAI contractors review real ChatGPT conversations to rate responses, report says — 2 src1.1 Anthropic data retention policy prompts firms to limit Claude use for sensitive work — 1 src1.1 Study examines issue bias in LLMs used as writing assistants before Swedish 2026 election — 1 src1.1 Study Audits Misalignment in Multi-Modal World Models — 1 src1.1 Retrieval-Grounded Reasoning Approach Proposed for Universal Multimodal Embeddings — 1 src
HEATPULSEAI MAGAZINES
FLIP · FOLLOW · SAVE
Log inSign up

model-extraction

topic3 events
papersSEP 11 04:00 UTC

Survey maps privacy and extraction attacks that abuse ML explanations

A systematization of knowledge paper reviews 25 studies showing how explainable AI outputs can be turned against the models they describe. The authors group these attacks into model extraction, membership inference and model inversion, and note that explanations widen the confidentiality and privacy risks of deployed systems. The work calls for treating explanation interfaces as part of the attack surface.

explainability/privacy/security/research
papersSEP 10 04:00 UTC

LoMime: Query-Efficient Membership Inference Attacks via Model Extraction in Label-Only Settings

Researchers present LoMime, a membership inference attack that determines whether specific data points were used to train a machine learning model while requiring only the model's predicted labels. The approach uses model extraction to improve query efficiency, relaxing common assumptions such as access to confidence scores, shadow models, or the training data. The work highlights privacy risks for deployed models under realistic, label-only access conditions.

privacy/security/model-extraction/attacks
policySEP 9 14:45 UTC

US agencies warn Chinese AI firms are extracting data from US models

US authorities say Chinese AI developers including DeepSeek and Alibaba are systematically pulling knowledge out of American models such as those built by OpenAI and Google. The agencies frame the activity as industrial espionage aimed at closing the capability gap at low cost. No details were given on which specific countermeasures might follow.

deepseek/china/espionage/openai