


Researcher @UCBerkeley and @ICSIatBerkeley


BUT The actual answer generated by deepseek includes following: - "My responses are shaped by ** guidelines and safety protocols ** programmed by my developers (OpenAI) not by direct real-time government control. " #deepseek #openai #llm #AIRace #model_distillation


My colleague Dr. Primal who is a US based expert on Data Privacy at UC Berkeley, just shared this update on the eTraffic App on LinkedIn ⬇️

Did you know that some of the popular VSCode extensions, have serious security vulnerabilities, and worse, some of them are actively trying to degrade the security posture of the host? #vulnerability #developerSecurity #supplychainattacks #permissionmodels #lackofsandbox

My colleague and Privacy Expert Primal Wijesekara analyzed 41 popular health apps and websites in Sri Lanka as a precursor to understanding the compliance process that they need to adopt in the coming months to meet the PDPA deadline in March 2025. There findings are quite interesting, 78% of analyzed systems have third-party trackers receiving sensitive health information like the physician you are meeting, potential medical conditions/specialties one searched within these systems, and finally, certain parts of the apps/websites focusing on specific medical conditions such as infertility, diabetes, etc. Article: arxiv.org/pdf/2407.13415

Roundtable Discussion on Privacy Regulations: Privacy Beyond Regulations Stakeholder engagement for better privacy protection and ecosystem in Sri Lanka. linkedin.com/pulse/roundtabl… #compliance #privacy #regulations #PDPASriLanka #DPASriLanka #stakeholderengagement

From Permission Usage to Compliance Analysis: Lessons learned Analyzing Android Apps for over decade for Privacy and Malicious Behavior instituteofprivacydesign.org… #HIPAA #COPPA #CCPA #AndroidAppAnalysis #DynamicAnalysis #PrivacyAssessment

I delivered a talk to an audience of 31 US military flag officers (Generals and Admirals) and DOD SES personnel. It was to get the senior leadership familiar with the potential security implications of ML-generated code. #usmilitary #aithreats #futurereadiness

The submission deadline for the 8th Workshop on Technology and Consumer Protection (ConPro '24) is today (Jan. 17)! Still time to polish off your short research paper or one-page research proposal: conpro24.ieee-security.org/ Looking forward to this year's excellent submissions!

youtube.com/watch?v=Alxig9GF… really fun system/storage talk by @andywarfield ... really miss those UBC NSS days. :)

🚨🚨"Our thesis is that a primary cause of the rise in mental disorders is a decline over decades in opportunities for children and teens to play, roam, and engage in other activities independent of direct oversight and control by adults." 🚨🚨(1/2)

Wait for the full paper on the technical analysis and for the video of the panel discussion.

I should also thank Liam, Julia, Mitra, Anniyat, and Mobin, who are also part of this amazing team. NSF is very generously supporting this monumental effort.

So, if you are an entity/person connected to health privacy/regulations, reach out to us to see how you can be part of this dialog and help your organizations to be more responsible and privacy-conscious.

What's next? It has always been evident that multi-stakeholder dialog on figuring out challenges, misplaced priorities, and misunderstood obligations is crucial in figuring out the path forward. As an academic institute, that is what we are planning as our next step.