Duality Technologies

The direction of travel for data and AI regulations is clear: privacy enhancing technologies (PETs) are a requirement. If not familiar with this segment of techs, then this may sound like work best left until it's a hard requirement. And that's where most people would be wrong, very wrong. By providing guardrails in place of data checkpoints, PETs accelerate data-driven growth and solve one of the main challenges for AI initiatives: data acquisition. Join our expert panelists to discuss the current state of regulations and discuss why regulators are putting so much time into recommending and publishing case studies using privacy enhancing technologies (PETs). Caroline Louveaux, Chief Data Privacy and Responsibility Officer, Mastercard Daniel Weitzner, MIT, 3Com, ex-Deputy CTO of US WH Paul Comerford UK ICO Ronen Cohen, VP Product Strategy & Regulatory, Duality Technologies #dataprivacy #nistAI #euaiact #regulations #dataprivacy #datasecurity #modelsecurity #secureAI

How do you maintain business growth and innovation plans with all these new data and AI regulations? Join us on 17Dec where we'll have expert regulators, technologists, and business leaders discussing approaches to moving confidently with data and AI initiatives while ensuring compliance with current and coming regulations. What's different with this round of regulations, is the detailed guidance provided by regulators, ushering in "the new way" to handle data moving forward; faster, more secure, and more valuable. Paul Comerford from the UK ICO will discuss regulations and the work done to help guide organizations forward. Caroline Louveaux Chief Privacy and Data Responsibility Officer at Mastercard will share her approach to AI initiatives amidst the changes. Daniel Weitzner of MIT and former US Whitehouse CTO, brings deep technical knowledge and experience with privacy techs as well as a valuable perspective on US standards and policies regarding privacy and security. Ronen Cohen's background in financial crime fighting and current work in the development and operationalization of advanced technologies will help illuminate key use cases, technologies, and expectations for those exploring options. Reserve your spot today: https://lnkd.in/eCgeGsX9 #dataprivacy #regulations #SecureAI #GDPR #NIST #datasecurity #modelsecurity #dataops #MLSecOps #federatedlearning #confidentialAI

The Adtech ecosystem involves a significant amount of personal information as it changes hands among collectors, processors, aggregators, and users of that data. With traditional means of using data, this creates a lot of risk. Must we surrender privacy for personalized services? Or can privacy techs minimize or eliminate the risks in this ecosystem while balancing the needs of businesses and customers? #dataprivacy #innovation #datadrivengrowth

Happy Regulatory Tuesday, Are new data and AI regulations truly at odds with innovation? Only if you're not paying attention. A closer look into the recommendations of regulators and experts show that the new ways to handle and use data will accelerate innovation: - gain data speed with guardrails that replace compliance checkpoints - unlock data not previously accessible - opportunities to link datasets for even greater value - protect model IP while training across disparate datasets Read the full article here: https://lnkd.in/eGusXWJG #regulations #dataprivacy #innovation #datasecurity #privacytechnologies #securedatasharing

Happy AI Friday! In the spirit of today's consumer "holiday", let's talk about AI in Retail. "Artificial intelligence is already transforming retail, with almost seven of 10 (69 percent) retailers posting increased revenue after adopting AI, a recent Nvidia study shows. That sets the stage for AI to have an even bigger presence this year as the holiday shopping season approaches." How? The uses range from better visual search to better product visuals created by AI, product discovery, tailored recommendations, and price optimization to customer-facing LLMs to help locate products and services and streamline decision making. Let's break these down: Visual search How often have you seen a furnishing you like, snap a picture, but struggle to find something similar? Retailers are working hard to make this an easy workflow where a user submits such a picture and the model returns relevant results to suit the style. Personalized Shopping LLMs can learn the styles and preferences of customers, but can also be immensely valuable in helping a customer locate what they need—all without human involvement. It's no wonder this tops the use cases for retailers as models are far more efficient and accurate than a human counterpart could be. Better service and more sales for less, making your Black Friday and Cyber Monday even more successful. Read more: https://lnkd.in/grrVed4d

MSPs with clients struggling to move AI or data initiatives forward given new data and AI regs? Here's a good starting point. Whenever new technologies come on to the scene, technical teams must figure out how to address new issues--often with in-house solutions while vendors work on market-ready solutions. - Replace compliance checkpoints with data guardrails - Allow your customers to securely train models on their data before making a decision - Enable your clients to monetize their data with AI providers, securely - Protect model IP while training on disparate datasets Read the full article below: https://lnkd.in/e6zqf5sy #MSP #channel #secureAI #datasecops #dataops #MLops #innovation

Happy Regulatory Tuesday! Today's post comes from a recent bill being proposed to boost public services and the UK economy. Health and social care secretary, The Rt. Hon. Wes Streeting MP says: 1️⃣"This Bill and our Ten Year Health Plan will ensure important data flows safely and securely through the NHS, freeing up staff time and speeding up patient care." 2️⃣ "I know people worry about Big Brother, which is why data will only be shared to the most relevant staff and anybody using data must comply with strict security protocols." But won't strict security protocols work against data-sharing goals like they have in the past? Yes, but only if we're ignoring the new technologies the UK Information Commissioner's Office (ICO) has been publishing case studies about. For instance, the UK ICO published its PETs Guide, showcasing exactly how new technologies can be used to simultaneously improve data protection data use and data sharing. They've already published case studies using technologies that reduce the overhead costs, time, and limitations compared to traditional technologies Read more here - https://lnkd.in/eu-EktYy #UKBillofUse #DataSharing #PrivacyEnhancingTechnologies #PublicServices #TechnologyTrends #FutureofAI

"More than 3 billion phone coordinates collected by a US data broker expose the detailed movements of US military and intelligence workers in Germany—and the Pentagon is powerless to stop it." Those are big numbers and certainly cause for concern. How can we monitor and protect key people without disclosing who it is we're monitoring? It's this type of question that lands right in the "wheelhouse" of platforms built upon privacy technologies. The SOAA published a whitepaper (attached) discussing how they will address this risk using a PETs-enabled platform. To read the full wired article explaining the context of the problem, check here: https://lnkd.in/eb9_dKni #dataprivacy #opsec #dataprotection #dataops #crossborderdata #datasharing

The public sector is marching towards the new way to handle data: faster, more secure, and more valuable. Our public sector advisory team now includes experts from both the US & UK: Mi6, UK Joint Forces Command, NHS, US Whitehouse, US Navy, NSA. * Public-to-public: increasing the speed and scale of data among agencies, including hi-low workflows, is key to improving operational effectiveness and program success. * Public-to-public (allies): could be tackling international cybercrime rings, collaborating in a pandemic responses, national defense, or for rare disease research. * Public-to-private: reduce time and money spent to maintain OpSec while leveraging PAI/CAI/3rd party data. Read the full press release here: https://lnkd.in/dwphDnP4 Here's what they have to say:

Happy AI Friday! Watch our Co-founder and CTO, Dr. Kurt Rohloff, provides a summary of milestones in both the standardization and market-readiness of homomorphic encryption. Who should care? As Kurt says, the events are like "speed dating" for users and vendors: 1. find the right stack and get community assistance in operationalizing it into your solution 2. find a vendor that already has a market-ready solution that makes it easy to fit into your workflows Watch the full interview: https://lnkd.in/dx_sS4q8