Fears of tech rigging in 2022

A recent revelation by an American activist group about how Facebook has in the past one year used its algorithms to spread misinformation to 3.8 billion viewers has opened doubts over the credibility of Kenya’s 2022 General Election.

The social media behemoth has in the past two elections allowed Cambridge Analytica to mine profile and activity data of millions of Kenyans that it used to influence voter decisions using Artificial Intelligence (AI).

The Sunday Nation has established that contrary to many reports that only one company was involved in the psychological manipulation of voters in Kenya, seven other companies worked alongside Cambridge Analytica to keep the Jubilee Party in power.

Using Natural Language Processing (NLP), a sub-branch of AI that was initially developed for research purposes at Cambridge University by psychology researcher Aleksandr Kogan, the companies would pair individual psychological profiles with personal data obtained from the Facebook pages of quiz-takers and their online friends.

Text analytics

Through the use of text analytics, anchored on the Ocean Model of Personality, the eight companies analysed both structured and unstructured data from supporters of the opposition presidential candidate and devised strategies to tilt the vote.

 “That helped the companies know several traits of voters such as their political inclination, their social interests, levels of trust in key politicians, voting intentions, their gullibility while identifying introverts and extroverts,” Timothy Oriedo, founder of local data science training firm Predictive Analytics Lab told Sunday Nation.

That data became the root of a psychometric approach to targeting users with political adverts tailored to their personalities.

Those who supported Mr Raila Odinga would therefore receive targeted political ads on their Facebook timelines, often short videos and graphics that painted the ODM party leader as a man not fit for State House.

Both mainstream media and Kenya’s blogging community were used to change how voters viewed Mr Odinga, and successfully achieved to get a swing vote that got President Uhuru Kenyatta elected.

Through various public relations firms, the narrative was well designed and sent to newsrooms and bloggers to keep the Jubilee presidential campaign vibrant and dominant over its competitor.

Dirty political tactics

A TV expose by British Channel 4 News in March 2018 showed Mark Turnbull, the managing director of Cambridge Analytica’s political arm, thumping his chest about their dirty political tactics during the 2013 and 2017 campaigns, admitting that they ran the entire show for the Jubilee party.

"We have rebranded the entire party twice, written the manifesto, done research, analysis and messaging,” he bragged, adding that the company also wrote all the speeches and controlled about every presidential campaign element of the party.

Denying the allegations, the company said that it employs social media platforms for placing paid advertisements and organic content just like any marketing agency.

“Influencer marketing and building grassroots networks on social media are both common activities for a modern political campaign," it said.

The company reportedly conducted two surveys of at least 47,000 participants and earned millions of dollars from in the process, in what underpins Kenya’s dalliance with modern Western imperialism that is using Big Data analytics.

 “Political data mining companies are interested in presidential elections because of the high stake in the global geopolitical arena and the money involved,” Mr Oriedo notes.

Buts since the closure of Cambridge Analytica in 2018, the remaining seven companies have changed their names, closed their websites and rebranded, and are still operating in Kenya awaiting the next election.

 “We cannot rule out the possibility of the same happening again in 2022. They are closely observing the political developments in the country and will approach the most promising ‘customer’ as time goes by,” said a political expert who begged for anonymity.

Facebook, which has in the recent past given endless statements about how it protects user data, going ahead to launch what it calls Data for Good Project, is regarded as one of the most untrustworthy social media companies, even as it stores data of over 3 billion users across its three platforms – Facebook, Instagram and WhatsApp.

It has previously been castigated for working with Cambridge Analytica to sway voters in the 2016 US presidential election and the Brexit referendum.

In March 2019, the company was under fire for data protection failures, which led to the exposure of 600 million passwords stored in plain text.

Interestingly, the passwords were accessible by up to 20,000 employees of the social network, which defended itself that it had resolved a "glitch" that had stored the passwords on its internal network.

But the recent report by US activist organization Avaaz shows that misinformation on the social site rose during the Covid-19 period, with only 16 percent of it carrying a warning label despite a statement by the company that it “applied warning labels to 98 million pieces of Covid-19 misinformation and removed seven million pieces of content that could lead to imminent harm.”

Avaaz campaign director Fadi Quran said Facebook’s algorithm is sabotaging efforts to provide reliable information.

"This infodemic will make the pandemic worse unless Facebook detoxifies its algorithm and provides corrections to everyone exposed to these viral lies."

To prevent the mining of user data from Facebook and other companies on smartphones, Apple has since moved to make it harder for any app to mine user data without their consent in iOS 14 iPhones.