Predictive analytics in communication: Maneuvering the ethical minefield

To say that predictive analytics is a driver of inequities in the society may be an overstatement. Or is it?

As technology progresses and data reigns supreme, on the one hand it is a boon to harvest data that paves the way for smart decision making, but on the other, it feeds on invasion of privacy, if not done within ethical guardrails. The data that fuels predictive analytics, in many ways, enables efficiency, maximizing the power of algorithms with historical data being used to give a peek into future trends. At the same time, it overlooks certain sections of society that are bereft of communication access and subsumes the voices of minorities.

There exists a fragile point of equilibrium in the dilemma of safeguarding privacy versus leveraging data for utility – that’s where ethics come in. To maintain the integrity of ethical standards in communication, a strong ethics framework is the guiding compass, irrespective of the juncture in the lifecycle – be it data collation, market analytics, or communications – all of which are building blocks for the output generated by predictive analytics.

Let’s explore some real-life instances:

• Netflix's opaque recommendation algorithm raised concerns about data privacy and manipulation. Even though, Netflix, by using big data and predictive analytics, has managed to clock a retention rate of 93%. In fact, actively listening to consumer preferences, may also be how they cracked the code to create films that have won Oscars.
  
  
• Uber's use of predictive analytics to manipulate pricing during peak hours led to accusations of unethical business practices, tarnishing the company's reputation.
  
  
• Like many other companies using AI recruitment systems that turn out to be gender/age/cultural background-biased, iTutor Group’s AI recruitment system automatically rejected over 200 qualified applicants based on age. This again, highlights concerns about algorithmic bias and discrimination.
  
  
• Facebook's Cambridge Analytica scandal demonstrated the ethical risks of data misuse, emphasizing the importance of ethical data handling practices.
  
  
• Predictive analysis carried out by Target resulted in showing ads to pregnant shoppers, teenagers included. This sparked controversy over privacy invasion and ethical boundaries.

It truly is a complex mesh of ethical implications woven into predictive analytics, that commands the relentless pursuit of consensual and secure data-driven insights.

Challenges and key considerations:

I appreciate brands using data science and as a communications professional, I too make use predictive analytics to understand customer needs and trends, and generative AI tools to make life simpler. However, to rely solely on technology does not cut it. The onus of validating facts and communicating transparently lies on us. Afterall, I can’t be the only one getting exasperated when a chatbot asks for a photographic proof of an item that never was delivered; or when ads pop up following a non-descript conversation you had “in private” – that begs the question, when did my IoT powered gadgets enter my circle of trust… why did it enter the private chat and never left? Am I not the admin of my life anymore?


In an all-tech world, sleek and smart,
Devices eavesdrop, play their part!
Ads pop up, with eerie precision,
Targeting with startling incision.

From kitchen chats to bedtime tales,
They're tuned in, setting sales.
For in this world of ads and noise,
Your privacy's among their favorite ploys!

So, what are some of the challenges that pervade the world of communications today?

• Adherence to DPDPA/GDPR/similar regulations
  Example: Google significantly changed their data collection, processing and communication practices to ensure compliance with GDPR.
  
  
• GenAI as a questionable source of information
  Example: “ChatGPT, if the accurate answer isn’t available, will make up or ‘hallucinate’ an answer, resulting in misinformation. The AI tools are only as good as the available data, and it’s going to replicate any biases or stereotypes generated by humans” said Staley, a member of PRSA’s Board of Ethical and Professional Standards (BEPS).
  
  
• Lack of fair representation
  Example: Upon realizing that in India, despite 80% of the farmers being women, there was hardly any representation of this on the web. DS Group took it upon themselves to create a repository of images that justified the data – this helps feed balanced and equitable information into predictive analytics tools in the longer run.
  
  
• Ideological polarization exacerbated by predictive analytics
  Example: As seen closer home or in the Biden-Trump elections, biased and skewed information was being showcased to individuals who took to the internet to learn more. Instead of being able to see objective information, people were fed stories concocted by AI tools and bots to strengthen ideological polarization.

Are communications professionals the gatekeepers promoting ethical practices?

Communications professionals are uniquely poised to gatekeep companies’ information. Ethical considerations in predictive analytics and AI are not only a legal imperative, but it also goes to affect consumers’ loyalty, perception of brands’ trustworthiness and credibility. This pronounces the role of PR agencies, CMOs, journalists and all communication professionals alike as torch bearers of ethical practices in predictive analytics. It is key to understand the nuanced line treading between unethical practices and serving consumer preferences.

• Human touch:
  With generative AI and machine learning creating content available to the public, there is rampant prevalence of deepfakes, false and biased information. The role of communications professionals then becomes increasingly pivotal to edit and review communications at every step. This goes a long way in balancing ideological polarization that follows biased/false information.
  
  
• Source of truth:
  Creating avenues to communicate transparently about the source of information, along with due consent to use consumer data is also key. Especially in the post-truth environment that engulfs us today, it’s important that for any communication intervention, we partner with data providers who adhere to ethical data collection and handling practices.

It is worth noting that as per a report by The Insight Partners, “the global market for predictive analytics solutions reached more than $12 billion in 2022 and in 2028, it might go up to $38 billion, with a CAGR of 20.4%.”⁴ Predictive analytics has the capacity to transform communication strategies, facilitating precise messaging and improved audience interaction. Nevertheless, ethical concerns stemming from predictive analytics necessitate meticulous attention and ethical conduct from both brands and communication experts. By placing importance on safeguarding privacy, reducing algorithmic bias, securing informed consent, promoting transparency, and demonstrating ethical leadership, companies can leverage predictive analytics effectively while preserving trust, credibility, and societal accountability. While grappling with the ethical intricacies of predictive analytics, brands not only can bolster their commercial achievements but also foster a more ethical and inclusive digital environment.

In the realm of data's silent hum,
Where predictive analytics deftly come,
As algorithms sift and sway,
Consent shadows mark the way.

Let truth and transparency be our guide,
In communication's ethical stride.
For in this dance of bytes and code,
Ethics and integrity must lead the road.