Business Ethics 101: Cambridge Analytica vs FB analysis from an AI perspective

#businessethics #facebook #cambridgeanalyticafacebooikscandal #AIethics #dataprotection #consumerrights

The famous data scandal between Cambridge Analytica and Facebook emerged since 2018 and ended up with 5 billion fine to Facebook's failure in user data protection. As a Facebook user by then, when I heard the News, I posted my comment immediately. It sounded interesting to me when they two were in fight because I lived in England before and then I was a Facebook user living in the Bay Area. Transitioning from Europe to the USA, I learned that privacy wasn't a thing digitally for most Americans.

I was a Facebook user only because my American friends use Facebook, and social media either wasn’t a thing at the places I previously lived such as Cambridge itself. We apparently used texts or phone calls. Mostly in person.

?

At Cambridge, United Kingdom, whenever digital devices go faulty, people would exclamate: “Ah- modern technology! ” This remark has not stopped until 2020. Then digitalisation gradually became a norm with the Covid pandemic. Before the concern of data protection and privacy exposure officially spread around the world, the scandal between the two firms did not successfully alert people of their personal data online.

?

As I wrote this article, by a search of the event timeline, I surprisingly found that the scandal was first exposed 5 years ago, exactly today. I know I’ve always been accurate and, I also reviewed Elizabeth Holmes Theranos Case on her birthday Feb 3, this year on my own Facebook. This type of thing has not been a coincidence to me since every coincidence has been a coincidence.

In my view, Cambridge Analytica as a business has performed well in every regard and served the political campaign the best algorithmic result in its legit capacity.

If we review the event timeline, in chronological order (citing from Wiki):

1.?Other advertising agencies have been implementing various forms of psychological targeting?for years and Facebook had patented a similar technology in 2012

2.?The data was collected through an app called "This Is Your Digital Life", developed by data scientist Aleksandr Kogan?and his company Global Science Research in 2013.

4.?Cambridge Analytica was allegedly hired as a consultant company for Leave.EU and the UK Independence Party during 2016, as an effort to convince people to support Brexit.

5.?Information about the data misuse was disclosed in 2018 by Christopher Wylie, a former Cambridge Analytica?employee, in interviews with The Guardian?and The New York Times. In response, Facebook apologized for their role in the data harvesting and their CEO Mark Zuckerberg?testified in front of Congress.

6. In July 2019, it was announced that Facebook was to be fined $5 billion by the Federal Trade Commission?due to its privacy violations.?In October 2019, Facebook agreed to pay a ￡500,000 fine to the UK Information Commissioner's Office?for exposing the data of its users to a "serious risk of harm".

The newest update to the lawsuit is from BBC: Meta settles Cambridge Analytica scandal case for $725 million. (Source: https://www.bbc.com/news/technology-64075067)

?

Continue citing from Wikipedia, "The official investigation by the UK Information Commissioner found that Cambridge Analytica was not involved "beyond some initial enquiries" and the regulator did not identify any "significant breaches" of data protection legislation or privacy or marketing regulations "which met the threshold for formal regulatory action".

Apparently, it is very naive to comment that Cambridge Analytica did not do well in its business. But if we stand in the social media developer's perspective to look at the whole picture, it is also easy to understand that networks are developed and data are easily handled off and connected to different parts of the world online. After all, it is you who put yourself online.

What was wrong is Facebook’s side. They did not protect users’ data. Instead, they were proudly announcing multiple third-party tools that facilitate the user's interests and

And -

"Less than five weeks after Facebook launched Open Graph API version 1.0 for developers, Zuckerberg wrote an op-ed for The Washington Post?in which he vowed to resolve users’ concerns about how their personal information was being managed." from CNBC.

He said: “We have also heard that some people don’t understand how their personal information is used and worry that it is shared in ways they don’t want. I’d like to clear that up now.”

?

After he cleared it up, in 2010, the scandal still happened, with a full explosion of data leakage, privacy exposure, to all of the political campaigning.

?

I definitely don’t believe politicians don’t want less votes, so I definitely believe Cambridge Analytica could be a great business. And if they and Facebook have had some sort of business agreement then Facebook should provide the data according to the corporate contract, and under consent of their user if it was written in their policy- it seems that most users consent to hand over their personal data anyway. Then Facebook, after all of the selling data to not only third-party, but also a political consulting firm, has also accused Cambridge Analytica of her “misuse” of data.

?

I mean, it's a political firm, their aim of existence is to provide politicians benefits, what else do you expect them to do?

?

So I laughed about it then back in 2018 and am still laughing about it now.

?

Then I researched about Cambridge Analytica -

“ Kaiser’s insights offer new clues. They provide hints on how Cambridge Analytica and its parent company SCL Group used AI to make its targeted advertising more effective; and they illustrate how AI has become an integral element of digital political campaigning. " (Source: https://www.politico.eu/newsletter/ai-decoded/politico-ai-decoded-how-cambridge-analytica-used-ai-no-google-didnt-call-for-a-ban-on-face-recognition-restricting-ai-exports/)

They are actually one of the earliest to use AI to psychoprofiling, in another word, microtargeting.

Actually, businesses could morally use users data to still do psychoprofiling with their consent, as long as the purpose is explicitly clear.

A positive example is what I found in "The Twitter Exoplorer" Project.

https://arxiv.org/pdf/2003.03599.pdf

?

This article morally explored Twitter’s user data and as a result, the research could be used for today’s political solving.

?

“ We collected data using the keyword "Brexit" about 10 days before the General Election in the UK in December 2019. We observe a polarized retweet network, where pro and anti-Brexiteers form two distinct clusters. “ at page 7 of the paper

?

From an AI perspective to comment on this "analytica" business,

right now, it is unclear if, psychoprofiling still works for the general public now.

Several years ago, American population probably were very ignorant of digital manipulation, but since the pandemic started and many other things happened, there tend to be more people who are aware of the privacy.

However, "awareness" is not skillfulness.

"Yet when presented with a series of true-or-false questions about how digital devices and services track users, most Americans struggled to answer them, according to a report published on Tuesday by the Annenberg School for Communication at the University of Pennsylvania."

(from https://www.nytimes.com/2023/02/07/technology/online-privacy-tracking-report.html )

To pertain online privacy is also very important, the important thing for consumers is,

would you like to be exposed online and get more ad recommendations and more personalised stuff?

If you are not contributing your data, why would you pay your internet fee and go online?

To be honest, there is certain utility in putting yourself completely online, and the real privacy is not what your Facebook profile suggests.

I found it comincal when THAT happened, the Cambridge Analytica Facebook Scandal, and then the Covid, the massive hacking, people are still using facebook- but with a different last names..

We talk about the fidelity of your information online. Of course, you can put a different profile picture and a different name on Facebook, but then this platform is just proven to be an anonymity-openess mixed platform. I think, Zuckerberg's initial aim to create such a social platform is to let people get online and know each other with a real name.

Yes, get online is fine, but getting online + knowing each other is difficult.

To know each other it has to be offline. Nothing can replace a real life interaction.

This is also why from an AI perspective, no matter how machines learn our data, it still can't speed up to our mind's race.

It is a race between machines and human minds, and AI is just innocently used as a tool - again, from a business perspective, it performed excellently seeing from the results of the election.

As for the ethical perspective, it is really, up to the general public to be educated that, next time not to be manipulated by the information online, that is why, Facebook sueing Cambridge Analytica would be proven a failure.

There left a "profound" question, should we socialise online?

Except third-parties who steal our data, the merchant, the person you are talking, and the trackers, cookies, everything exposes you, if you store your ID's picture on the phone, it WILL be exposed, in some form eventually handed to the hackers.

The cyberspace isn't a moral space.

The AI unfortunately also lives in the cyberspace. It is immoral to accuse AI of anything because she does not create herself.

The cyberspace is also a fun space, i.e: the project twitter explorer as illustrated above.

Twitter is another topic of social media use and I'm looking forward to seeing how Elon is going to make twitter-space great!

?To summarise, business ethics in the cyberspace means,

?

23:46, 3/17/2023 update,

He took my opinon and will implement the solution on Twitter AGAIN ( just 2 hours ago)!

We were JUST talking about public opinion manipulation!

I am so happy!!!!! Thanks again and again, Elon!

The AI/ML solution was suggested on my Linkedin previously here , on message automation which is executed and controlled by machine learning.