Digital Personal Data Protection Bill of India | Who benefits when AI speaks?
CyberInsights #105 - India's controversial DPDP || AI trustworthiness in answers
I missed the last weekly newsletter - travelling. I was in Malaysia, meeting the nice cyber insurance folks there and of course, enjoying the local food. I spoke about cyber insurance over some delicious Chicken Satay in KL…
Well, coming back to cybersecurity, lots of things this week.
Indian Parliament passes controversial Data Privacy bill
Experts say that it gives the government and private entities wide access to PII and shields transparency of the government actions.
Data is the new oil. Everyone wants to control oil. Hence, everyone wants to control data.
Download the latest version of the Digital Personal Data Protection Bill from this website.
There is quite a bit of controversy on the bill. The bill also amends another law titled ‘The Right to Information Act’ and dilutes it. Read some of the analysis:
Medianama - a very comprehensive resource. Includes previous drafts and comments and timelines.
Take Action:
Read the bill. It will soon be a law. As a citizen, check if your rights to privacy are being met and write to the government if they are not.
As an infosec professional, analyse what needs to be done to implement the law in your organisation. Follow me on LinkedIn or watch this space for a webinar on the same along with cyber lawyers.
Does Generative AI serve it’s masters?
Is generative AI prevented from speaking ill about its creators?
Well, Duh!! What did you expect? Read the image above. When I asked ChatGPT about Microsoft, it shied away from telling me that Microsoft was in fact called out for being a monopoly.
I asked ChatGPT if it has to answer positive things about its creators. This is what it had to say:
Clearly, if it had used public information unto September 2021, it would have read the news in 2014. In fact, Time had this in November 2014:
So, you know that generative AI serves its masters. Read this article by Bruce Schneier. He questions how the AI has been trained and what it is being fed. Can you inherently trust AI?
Take Action:
By now, you must be a part of an organisation that has at least 1 generative AI project in progress. Do an assessment of the same and figure out if your AI is free from biases and hallucinations.
As a user of generative AI - try to understand the biases that have deliberately been trained in the AI before using it willy nilly.