Seamus Blackmore is a partner and product leader at KPMG in Canada Lighthouse, the firm’s emerging technology practice. Ven Adamov is a partner and data and analytics leader in the firm’s risk consulting practice.
Generative artificial intelligence was introduced to the public barely one year ago, and it’s rapidly transformed engineering departments, boardrooms and kitchen tables across the country, with one in five Canadians using the technology at work or school to do research, generate ideas and create presentations, according to KPMG in Canada’s Generative AI Adoption Index.
While AI has been part of people’s lives for many years (think about your streaming service recommendations, predictive internet search engine results or the “for you” tab on your social media feed), the breakthroughs in generative AI – a type of AI that generates instant content on demand – are happening at an exponentially faster rate. Consider it a newer version of “Moore’s Law,” which dictates that the number of transistors on a microchip doubles every two years. Generative AI has turned Moore’s Law on its head, unleashing innovation at a pace previously unimaginable. Understandably, this is causing unease among business leaders and citizens alike.
That incredible pace of change means that what we think of as generative AI today will look very different a year from now. Already, the concept of the ‘large language model’ (the text generation that underpins generative AI) is getting outdated as more models become ‘multi modal’. For example, only a few months ago, users were generating content with simple text prompts. Now, they can ask generative AI platforms questions using their voice, and they’ll get answers in writing, audio, image, video or even 3D.
Business leaders need to be thinking about generative AI’s disruptive capabilities, how to navigate the challenges of this powerful technology, how to implement it responsibly and how to upskill talent for it. Every executive should have a baseline understanding of how their department could be disrupted by generative AI so that they can ride the wave instead of being crushed by it.
More than one-third (37 per cent) of Canadian businesses are experimenting with generative AI to see how it can improve their operations, compared with 65 per cent in the U.S. So even if you’re not thinking about generative AI, your competitors and market leaders might be.
The same goes for your employees – they might already be using generative AI tools. Our research shows nearly one quarter of users are putting sensitive company information into generative AI platforms such as ChatGPT, and less than half consistently check the accuracy of content generated by AI systems. Getting ahead of this will help reduce the risk of privacy and cybersecurity breaches, as well as intellectual property and financial damage.
It starts with data
Protecting company data is a crucial step in any organization’s action plan on generative AI, and employee awareness and education should be at the heart of this. Spelling out the “do’s” and “don’ts” with a comprehensive employee usage policy is a good place to start. Offering generative AI training and upskilling for employees is essential.
Encouraging a culture of experimentation and transparency with generative AI tools in a controlled environment could help your organization discover new use cases for generative AI. For example, many companies have held “prompt-a-thons,” where employees are challenged to generate new ideas and prompts to find solutions to common business problems. Exercises like this are great ways to spur innovation without compromising company data.
Staying up to date on evolving legal and regulatory changes is also something that all business leaders will need to do if they want to implement generative AI responsibly across their organizations. The federal government recently announced a voluntary code of conduct for companies developing and using generative AI, and many firms have already signed on to it. Adopting guardrails similar to what the federal government has laid out, can help organizations earn trust from consumers and citizens and demonstrates responsible integration of generative AI. Your guardrails should include setting up cross-departmental governance structures, employee awareness, education and training, clear and comprehensive policies on responsible usage and robust security and data privacy controls. With carefully and thoughtfully designed guardrails you’ll be better positioned to stay ahead of new and evolving regulations and help future-proof your strategy.
Building in-house
If you’re willing to go one step further to maintain privacy and security of your company’s data, your organization could also consider a private generative AI platform that’s trained on your proprietary data and can only be accessed by your employees. Many of Canada’s largest organizations are already using their own generative AI systems to manage inventory, schedule staff, improve supply chains, detect operational issues, generate marketing content and design 3D renderings. A private model requires commitment and investment to keep pace with new and evolving developments, but it’s customizable, reliable and it adds an extra layer of data security. Building it requires flexibility, agility, security, ethics, transparency, trust and reliability.
If this sounds like the right solution for your organization, consider the following:
- Explore the myriad technology vendors, services and solutions available to help build a model that’s in line with your company’s values and vision, and consider using or layering a few different solutions, as it’s not a “one size fits all” approach.
- Ensure the data used to train your model is accurate, recent, relevant and unbiased and your algorithms can be easily understood, accessed and explained by internal and external stakeholders.
- Build reliability into your generative AI model by reducing the frequency and likelihood of false or misleading content.
- Assign accountability for generative AI at the organizational, C-suite and board level.
- Implement strong cybersecurity and privacy controls.
- Train and empower your people to use it effectively and responsibly.
Exploring and understanding how generative AI can benefit your organization is the first step toward becoming a more productive and innovative company. But in the era of generative AI, anything and everything could change, which means that the most effective business leaders will be the ones focusing on strategies that work in times of unpredictability. This means making shorter-term plans, building in flexibility and making work forces more adaptable, investing in research and development, upskilling key decision makers and implementing agile controls and guardrails.
This column is part of Globe Careers’ Leadership Lab series, where executives and experts share their views and advice about the world of work. Find all Leadership Lab stories at tgam.ca/leadershiplab and guidelines for how to contribute to the column here.