Synthetic Data: The Future of AI and Privacy

most scientists think data privacy of their research is just being careful who they share the data with. it's a trust model. what they don't realize yet is OpenMatter enables trustless and verifiable collaboration on sensitive data.

Earlier this month, the Court of Justice of the European Union (CJEU) clarified something that could reshape how organizations use privacy-enhancing technologies (PETs). If a party cannot reasonably re-identify individuals in a dataset, even if it’s pseudonymized, that data may be treated as effectively anonymous. That distinction matters. It means technologies like Confidential Computing and Fully Homomorphic Encryption aren’t just technical safeguards they could legally transform how organizations handle regulated data. For cross-border research, financial analysis, or AI training, this ruling could reduce friction, speed up compliance reviews, and finally bring legal recognition to the promise of privacy-enhancing technologies (PETs). This is a topic we will be diving deeper into with a webinar from one of the leading experts in the field - stay tuned.

Earlier today, I participated in the Nordic Privacy & Innovation Summit in Stockholm, and the sessions offered a sharp snapshot of where Europe stands on privacy, AI and digital resilience. And of course the leaked Digital Omnibus was the big elephant in the room. Peter Craddock argued for a more pragmatic regulatory mindset - using the example of Transfer Impact Assessments to illustrate the enormous effort for limited real-world benefit, and how today’s rules are increasingly hampering needed European AI model training activities. Shervin Razani made an important assertion and distinction that AI-risk and regulatory risks are two very different creatures. Regulation, if mis-calibrated, can slow or even stall innovation long before the technology itself poses any meaningful threat. The excellent Evelina Anttila delivered the helicopter perspective. Her point was unmistakable. The next major crisis could be digital, leaving society offline for weeks. Trust, resilience and (AI) sovereignty will decide who copes and who doesn’t. She also noted how aggressively California is moving on AI regulation, backed by investors with far greater risk appetite, pulling European talent and companies westward. And as always, it was a pleasure to catch up with Malin, Birgitte, Kristín 😊 Time well spent, and plenty of ideas to bring back into our work with clients navigating data protection, cybersecurity and AI governance in 2025. Big thanks to Mannheimer Swartling for arranging the event, and kudos for drawing such an impressive turnout of privacy professionals.

AI and Data Governance: A Smarter Approach to Privacy Compliance As AI transforms how we manage and protect data, strong governance frameworks are key to staying compliant and building trust. Discover how to align innovation with responsible data use. 👇 https://lnkd.in/dVJFTn8i

The Ultimate Privacy Showdown: De-identified vs. Synthetic Data | New AI-Powered Podcast Episode (Epi 32) 🎙️ The defining challenge in data governance is here: Can De-identified Data still secure sensitive records against modern AI-driven threats? Or is Synthetic Data the necessary evolution? In Episode 32 of our AI-powered podcast, we conduct a deep dive into these two core strategies for protecting health information: ✅ De-identified Data: High fidelity, but persistent re-identification risk. ✅ Synthetic Data: Unbreakable privacy by design, but requires crucial utility validation. We unpack the trade-offs and explain why adopting a hybrid, fit-for-purpose model is now essential for ethical and regulatory compliance in the age of Generative AI. This is vital insight for data leaders, executives, and policymakers. 🎧 Tune In Now (Epi 32): YouTube (Video): https://lnkd.in/e3hrVdFV Spotify (Audio): https://lnkd.in/eMUdGH5z #SyntheticData #HealthTech #DataPrivacy #AIinHealthcare #DataGovernance #Epi32 #SeidData #healthepi #WHO

The Ultimate Privacy Showdown: De-identified vs. Synthetic Data | New AI-Powered Podcast Episode (Epi 32) 🎙️ The defining challenge in data governance is here: Can De-identified Data still secure sensitive records against modern AI-driven threats? Or is Synthetic Data the necessary evolution? In Episode 32 of our AI-powered podcast, we conduct a deep dive into these two core strategies for protecting health information: ✅ De-identified Data: High fidelity, but persistent re-identification risk. ✅ Synthetic Data: Unbreakable privacy by design, but requires crucial utility validation. We unpack the trade-offs and explain why adopting a hybrid, fit-for-purpose model is now essential for ethical and regulatory compliance in the age of Generative AI. This is a vital insight for data leaders, executives, and policymakers. 🎧 Tune In Now (Epi 32): YouTube (Video): https://lnkd.in/e3hrVdFV Spotify (Audio): https://lnkd.in/eMUdGH5z #SyntheticData #HealthTech #DataPrivacy #AIinHealthcare #DataGovernance #Epi32 #SeidData #healthepi #WHO

The Ultimate Privacy Showdown: De-identified vs. Synthetic Data | New AI-Powered Podcast Episode (Epi 32) 🎙️ The defining challenge in data governance is here: Can De-identified Data still secure sensitive records against modern AI-driven threats? Or is Synthetic Data the necessary evolution? In Episode 32 of our AI-powered podcast, we conduct a deep dive into these two core strategies for protecting health information: ✅ De-identified Data: High fidelity, but persistent re-identification risk. ✅ Synthetic Data: Unbreakable privacy by design, but requires crucial utility validation. We unpack the trade-offs and explain why adopting a hybrid, fit-for-purpose model is now essential for ethical and regulatory compliance in the age of Generative AI. This is a vital insight for data leaders, executives, and policymakers. 🎧 Tune In Now (Epi 32): YouTube (Video): https://lnkd.in/e3hrVdFV Spotify (Audio): https://lnkd.in/eMUdGH5z #SyntheticData #HealthTech #DataPrivacy #AIinHealthcare #DataGovernance #Epi32 #SeidData #healthepi #WHO

20th Annual Data Protection Practical Compliance Conference 2025 https://lnkd.in/enWppjG2 Speaker Highlight Growing Pains: EU Age Verification Hannah Perry, Partner, Mason Hayes & Curran LLP As digital platforms face increasing pressure to protect minors online, age verification has emerged as a critical, and often controversial, challenge for organisations. This session explores the complex intersection of privacy, safety, and compliance, addressing: ▪️ The paradox of protecting young users while preserving privacy ▪️ Technical, ethical, and legal challenges in age assurance ▪️ Emerging solutions including privacy-preserving AI and zero-knowledge proofs ▪️ Regulatory developments shaping age verification across the EU 6th - 7th November 2025 Dublin City Centre https://lnkd.in/enWppjG2 #PDPConf2025IRE #DataProtection

Files Sometimes, personal data isn’t stored in databases or organized structures but in a pile of files somewhere on the server. When a user requests to be forgotten, we must keep track of all the files they uploaded, which may contain personal data. The most obvious example is CVs, but it could also include conference submissions with names and emails, maybe pictures (especially ones containing personal documents), etc. Tip: It is always a good idea to have a path to the files somewhere in the database and a flag that signals whether it contains personal data or not. -------------------------------------------------------- If you are a legal professional interested in exploring privacy technology further, check out my course (link in the first comment 👇), or follow me for regular online workshops I host on various topics like Privacy in AI, Cookies, Privacy in HTTP, and more.

🔒 The future of analytics is privacy-focused and cookie-less. 📊 Discover how data science is evolving to protect user privacy while maintaining accuracy and insights.   💡 Learn how ethical data practices are redefining digital intelligence.  𝐋𝐞𝐚𝐫𝐧 𝐌𝐨𝐫𝐞: https://lnkd.in/gN4r6zPd   #DataPrivacy #EthicalAI #DataScience #GlobalTechCouncil