This week, I was invited to join ATxSummit in Singapore. You can learn more through my review.
If you enjoy this newsletter, please become our paid subscriber to help this keep going.
Highlights
India: CCTV security rules.AI: Secure by Design needed.Adidas: Third-party data breach.LexisNexis: 364K+ records exposed.Singapore: AI Governance Summit.Special!
Deep Dive
1. India's CCTV Crackdown Reuters
New rules since April 9th.
April 3rd meeting in Dehli.
17 manufactures requested delay.
Gov denied delay request.
More testing labs anticipated.
Headsup: US, UK, Australi imposed restriction.
2. Design Principles for AI Dark Reading
Reactive approach fails.
Data poisoning, prompt injection, and more.
Integrated into the AI life cycle.
OWASP Top 10 for LLMs.
Data prep, model training and test, deployment.
Headsup: Follow our review to get update from the latest news.
3. Adidas 3rd Vendor Breach Bleeping Computer
Names, email, phone, BoD, addresses.
Customer service provider compromised.
Ongoing investigation underway.
Earlier this month, disclosed breach 2024.
Included Turkey and South Korea 2024.
Headsup: Your third party security is also important.
4. Massive Data Broker Breach TechCrunch
LexisNexis: 364,000+ records exposed.
Breach occurred Dec 25, 2024.
Data includes SSNs, drivers licenses.
GitHub account compromised.
No ransom demand reported.
Headsup: Assess data broker relationships and potential exposure risks immediately.
5.Testing Real World AI system Global Assurance Pilot
17 org from health, bank, IT and more.
16 testing AI specialists
Simulation testing is highlighted.
Synthetic data for adeversarial testing.
Agentic AI is deployed as a judge.
Headsup: Check out their starter kit to test AI system.
Notice: Y2 GenAI Safety and Security is on GumRoad and Amazon with paperback.
Share this post