What Meta, Amazon and YouTube taught us about AI

We’re in the midst of an AI space race. And organisations across all industries are pushing boundaries at an unprecedented pace. From enhancing customer experiences to streamlining operations, AI promises to revolutionise the way we live and work.  

However, as these technologies become more ingrained in our daily lives, ethical missteps are proving costly. And the cost isn’t just financial, it damages public trust and reputation too. 

Meta recently found itself at the centre of a privacy storm to personalise user experiences. It integrated facial recognition technology into its Facebook service to automatically tag individuals in photos. What seemed like a seamless enhancement quickly spiralled into controversy when users discovered that their images were being analysed and tagged without explicit consent.  

Privacy arguments erupted, highlighting the potential for misuse and the erosion of personal data security. The backlash was swift and severe: user engagement temporarily plummeted, regulatory bodies around the world launched investigations, and the company’s reputation took a significant hit. Meta quickly rowed back, underscoring a fundamental flaw in deploying AI technologies without landing on a robust ethical framework. 

Some flaws are unintentional but are still ethically harmful. In 2018, Amazon developed an AI-driven hiring tool designed to streamline its recruitment process by analysing CVs and identifying top candidates. It would act as a critical co-pilot for the company’s People team. Only later, the tool was found to show significant biases against applicants who identified as women. This bias stemmed from the training data, which consisted of historical CVs submitted to Amazon over a ten-year period, mostly from candidates who identified as men. As a result, the AI system learned to favour candidates that identify as men by penalising applications that included terms commonly used by people that identify as women. 

And new, sinister, use cases of AI are presenting ethical dilemmas too. YouTube has faced numerous challenges in managing deepfake content. The platform has been used to upload hyper-realistic videos that mimic real individuals, spreading misinformation, scamming unsuspecting victims, manipulating public opinion, amplifying damaging fake news, and sometimes even defaming public figures. 

Amid these high-profile failures, a common thread emerges: the absence of a pragmatic approach to AI deployment. This is where the principles of our Pragmatic AI approach come into focus.  

Pragmatic AI advocates for a balanced and measured approach that prioritises simplicity, affordability, enablement, and measurable outcomes. Instead of diving deep into novel or innovative trends, the pragmatic approach emphasises use-case-driven innovation, measurable outcomes, and long-term sustainable growth. 

By anticipating ethical grey-areas and misuse, organisations can implement preventive measures. It means mitigating embarrassing headlines and hopefully swerving adverse effects while still leveraging the benefits of AI. 

These lessons highlight the critical importance of embedding ethical considerations into every stage of AI development and deployment. Pragmatic AI is not just a theoretical framework; it is a practical guide for navigating the ethical landscape of modern technology. By focusing on clear objectives, iterative development, and stakeholder engagement, organisations can avoid the pitfalls of overambition and costly failures. 

As AI continues to evolve, the stakes only get higher. Public trust is fragile, and once broken is challenging to rebuild. In the end, the promise of AI lies not just in its ability to innovate, but in its capacity to do so responsibly. Embracing Pragmatic AI ensures that as the boundaries of what is possible is pushed, it is done so with a steadfast commitment to ethics, integrity, and foresight. 

Maybe you’re wondering how Pragmatic AI can shepherd your organisation towards an ethical and successful AI deployment? Download our Pragmatic AI White Paper to learn about strategies that prioritise simplicity, affordability, enablement, and measurable outcomes, ensuring initiatives are both effective and ethically sound.