For LLMs, IBM’s NorthPole chip overcomes the tradeoff between speed and efficiencyResearchPeter Hess26 Sep 2024AI HardwareExploratory ScienceGenerative AISemiconductors
DARPA and IBM are ensuring that anyone can protect their AI systems from hackersNewsMike Murphy07 Feb 2024AIData and AI Security
An AI model trained on data that looks real but won’t leak personal informationResearchKim Martineau12 Dec 2023AIAI PrivacyData and AI SecurityFinanceFoundation Models
An open-source toolkit for debugging AI models of all data typesTechnical noteKevin Eykholt and Taesung Lee08 Sep 2023Adversarial Robustness and PrivacyAI TestingData and AI Security
Simplifying cloud security policies with AITechnical noteJulian Stephen and Shriti Priya13 Jul 2023Cloud SecurityData and AI SecuritySecurity
Protect yourself from bad neighbors in the cloudTechnical noteMichael V. Le, Salman Ahmed, Dan Williams, and Hani Jamjoom07 Jul 2023Data and AI SecuritySecurity
AI diffusion models can be tricked into generating manipulated imagesNewsKim Martineau05 Jun 2023AIAI TestingData and AI SecurityFoundation ModelsGenerative AISecurity
Saška Mojsilović wants to channel AI for good. She may also make you rethink sour cabbageNewsKim Martineau09 Feb 2023AI TransparencyData and AI SecurityExplainable AIFairness, Accountability, TransparencyMachine Learning
What is synthetic data?ExplainerKim Martineau and Rogerio Feris08 Feb 2023AIData and AI SecurityMachine Learning
Improving the efficiency of adversarial robustness defensesTechnical noteKevin Eykholt20 Jan 2023AIData and AI Security
Five ways IBM is using synthetic data to improve AI modelsResearchKim Martineau02 Nov 2022AIData and AI SecurityMachine Learning
A new way to generate synthetic data for pretraining computer vision modelsNewsKim Martineau23 Jun 2022AIData and AI SecurityMachine Learning
Securing AI systems with adversarial robustnessDeep DivePin-Yu Chen15 Dec 20218 minute readAdversarial Robustness and PrivacyAIData and AI Security
Researchers develop defenses against deep learning hack attacksReleaseAmbrish Rawat, Killian Levacher, and Mathieu Sinn05 Aug 20217 minute readAdversarial Robustness and PrivacyData and AI SecurityGenerative AISecurityTrustworthy AI
AI goes anonymous during training to boost privacy protectionReleaseAbigail Goldsteen26 Jan 20214 minute readAdversarial Robustness and PrivacyData and AI SecurityMachine LearningSecurity