Page 63 - UKZN Proceedings of the Conference Report

assisting in curbing crime and ensuring public safety through the use of purpose-built surveillance equipment such as CCTVs, among others. Nevertheless, several scholars, such as Gershgorn (2018, p. 1) and Brian Brackeen, the founder of Kairos an innovative FRT company, have pointed out that FRT is still not mature enough to be applied on a large scale within the current legal and ethical frameworks. Issues like variations of light and locomotion can compromise the performance of FRT systems, thus introducing mistakes and prejudices that require removal (Campbell 2018, p. 2; White 2016, p. 1).
Another advancement in the use of technology for law enforcement is the concept of predictive policing, which is the practice of using data to predict the likelihood of criminal activity occurring. As Heaven (2020) puts it, such a model is armed with historical data with a focus on determining places that are most likely to experience crimes, facilitating pre-emptive actions by law enforcement. Whereas reactive policing relies on acting after crimes have been committed, predictive and intelligence-led policing involves collecting information about certain geographical areas and the population residing in them with a view to committing crime – and then intervening before the crime is committed (Strom 2017; OSCE 2017). Nonetheless, it is important to note that these systems are based on a wide range of data, which can be very stressful to the officers and lead to neglecting some crucial evidence (Fayyad and Uthurusamy 2002). This calls for appropriate systems for management of information here and also the use of AI.
Law enforcement agencies stand to benefit by incorporating AI in their work, as this reduces the chance of mistakes and biases. Research by Christie (2021) is in agreement that AI will be helpful in law enforcement, in that it will enhance the objectivity and consistency of decision- making. As reported by Deloitte (2021), even at its infancy AI has enabled police services to cut down on crime by 30–40% and enhance emergency calls responses by 20– 35%. At the same time, there are still worries regarding bias in the algorithms, especially when such systems are implemented, given that they could encourage negative tendencies reserved for those in power, especially in underclass populations (Guariglia 2022).
Body-worn cameras, drones, CCTV systems, and similar technologies are now essential components of policing. They help provide evidence and information at the time that the criminal activities are taking place, and help to hold the police accountable for their conduct. For instance, in the United States of America the use of body-worn cameras has improved police accountability and led the public to trust the police more (Stro 2017, pp. 4-13). In the same vein, service CCTV systems are installed in various locations and equipped with software which monitors the
camera coverage in real time and raises the alarm when certain behaviours are detected (Kwet 2019, p. 4). The Khayelitsha Report (2014, Recommendation 18, para 87) also recommended increasing surveillance, especially the use of CCTV in key areas such as transportation centres and educational institutions, to promote safety and control crime. Controlled aggressiveness of humans can also be ‘reinforced’ by using AI surveillance techniques, which can support proactive prevention strategies in policing rather than continue to allow reactive measures.
All in all, creativity in the use of technology, especially AI, in policy making would go a long way in solving most of the issues that face law enforcement in South Africa. Technology has great benefits for a police force that should not be dismissed, no matter the literature related to privacy, bias, and accuracy. If these technologies are to be implemented, it is paramount to the public image of SAPS that these technologies are adopted and implemented in a lawful and morally acceptable manner to promote trust in policing efforts.
Employment and economic empowerment
From one perspective, South Africa’s AI ecosystem has the power to change the country’s labour market while also mitigating the negative repercussions it might have on employment and economic growth. The development of AI is causing a disruptive change across all sectors of the economy, providing both advantages and challenges to the labour force. Whereas most pundits believe that AI will lead to improvements in productivity and effectiveness, there are fears that this will come at the expense of many jobs, especially those which involve performing similar tasks that are manual in nature. According to Gilbert (2023), AI has three main primary impacts on the workforce: job displacement, job augmentation, and job creation. This illustrates the ambivalence of the impact of AI on the employment of people – it dislodges existing paradigms of employment, but also creates new opportunities. In the light of these changes, the introduction of AI technologies must be moderated by policies to ensure economic advancement and job opportunities without leaving anyone behind. Adoption of AI innovations should take place along with education and reschooling so that workers have the right skills.
The capabilities of AI are most visible when talking of the emergence of new occupations and industries. According to the World Economic Forum (2023a), by 2027 fast-growing job categories will include specialists in AI and machine learning, business intelligence analysts, and sustainability analysts. These job titles are indicative of the rising need for advanced skills that can foster the development of AI and its application within different fields. In addition, it has been said that AI will be responsible for the emergence of new jobs, such as prompt engineers, AI modellers, and
Proceedings of the conference on Public innovation, develoPment and sustainability | 61