With the advent of accessible generative artificial intelligence (gen AI) tools such as ChatGPT, more and more organisations are experimenting with the technology and weaving it into their day-to-day processes.
As gen AI advances in leaps and strides, it is sure to become as much a part of our working lives as productivity software or the Internet, writes Dean Richards, CEO of Ricoh South Africa.
The South African Generative AI Roadmap 2024 from World Wide Worx shows that 90% of respondents from large South African enterprises currently use or have concrete plans to integrate gen AI within their organisations. Nearly all (95,6%) agree gen AI will have a positive potential impact on productivity, competitiveness, organisation of work, logistics, sales, and customer service.
There is little doubt about gen AI’s potential to turbocharge productivity and enhance user experiences by reducing the need for workers to carry out repetitive, laborious tasks. But organisations should not underestimate the importance of organisational culture and the human factor in adopting AI.
Ricoh’s research in Europe found that many organisations are lagging in providing employees with training and advice around the use and deployment of emerging technologies in the workplace. Our research found that only 18% of employers have implemented risk management measures around AI.
However, gen AI tools can expose businesses to a multitude of copyright, privacy and compliance risks if they are deployed without appropriate guardrails. Low quality training data and poorly written prompts can cause large language models to produce biased, inaccurate, or outright false outputs.
Humans in the loop
Relying on AI tools without proper human validation can lead to the unchecked spread of misinformation – internally, as well as externally. At a minimum, businesses must ensure that review processes are in place to catch erroneous AI outputs before they cause harm. But reviewing outputs is only one piece of the puzzle.
Without guidance on how to use AI appropriately, employees may inadvertently use the tools in ways that could compromise security, privacy, compliance, or result in biased and inaccurate outputs. Comprehensive governance of AI usage requires establishing policies around what types of data can be used for training models and how AI generated content is used.
Policies around how synthetic data is used in the workplace, as well as training employees in the art of prompt engineering, is essential to prevent biases in outputs, which could lead to inaccurate or poor-quality work. It is also advisable to involve legal teams from ideation to avoid risks later down the line and accelerate AI maturity.
Foster a culture of experimentation
Companies should also foster an environment where employees feel comfortable reporting issues or unintended consequences, so that corrective actions can be taken when needed. AI technology is in its infancy. By cultivating a culture of learning, companies can grow their collective knowledge of AI and help employees to continuously improve their skills and productivity with the new tech.
Businesses must take control of their AI future through workforce education, ethical guidelines and help from experienced partners. With preparation, companies can unlock AI’s potential to empower employees and transform operations–combining people and technology to drive productivity, innovation and competitive advantage.
Organisations that act decisively will thrive in the age of AI.