For business leaders, there are plenty of reasons to be excited about generative AI, starting with its power and ease of use. But, as with any emerging technology, there are also potential new risks. Some of these risks may come from your company’s use, others from malicious actors.
To manage both kinds of risks and harness generative AI’s power to drive sustained outcomes and build trust, you’ll need responsible AI. What exactly is responsible AI? It is a methodology designed to enable AI’s trusted, ethical use. It has always been important, but it has become crucial in the dawning era of generative AI.
With plain language commands, you can direct generative AI to create software code, data analysis, texts, videos, human-like voices, metaverse spaces and more. As the months pass, this power is expected to transform functions, business models and industries. Already, generative AI is being used by more people, not just data scientists, to deliver value at scale. It’s boosting productivity, supporting human decision-making and reducing costs. Use cases that are scaling up today include:
Generative AI is far from perfect. It tries to predict the right response to a prompt or request, based on its algorithms and data sets. Its outputs can be highly useful, but they may resemble “first drafts.” You’ll likely need to verify these drafts’ output and analyze their quality, then modify them appropriately. At times, generative AI can also produce irrelevant, inaccurate, offensive or legally problematic outputs — in part because users may not give generative AI models the right prompts and in part because these models are by nature creative. New cyber risks are also emerging, including the ability for malicious actors to use these technologies to generate deep fakes and facilitate other cyber attacks at scale. Because more people than ever can use generative AI, some of these risks may also become more widespread: many generative AI users within your organization may be unfamiliar with how AI models work and how they should be used.
Responsible AI can help to manage these risks and others too. It can grow trust in all the AI that you buy, build and use — including generative AI. When well deployed, it addresses both application-level risks, such as lapses in performance, security and control, and enterprise and national-level risks, such as compliance, potential hits to the balance sheet and brand, job displacement and misinformation. But to be effective, responsible AI should be a fundamental part of your AI strategy.
Our approach to responsible AI, PwC’s Responsible AI, delivers trust-by-design through the entire AI life cycle with frameworks, templates and code-based assets. It’s relevant to executives at every level: The CEO and board set the strategy, with special attention to public policy developments and to corporate purpose and values. Chief risk and compliance officers are in charge of control, including governance, compliance and risk management. Chief information and information security officers take the lead on responsible practices, such as cybersecurity, privacy and performance. Data scientists and business domain specialists apply responsible core practices as they develop use cases, formulate problems and prompts and validate and monitor outputs.
As your business begins exploring generative AI use cases and how to apply responsible AI, here are some key risks to be mindful of:
If you currently have a strong responsible AI program, your governance efforts have probably already flagged many of generative AI’s new challenges. Still, there are key areas to pay close attention to and key steps to consider in applying responsible AI to a tech environment that’s quickly evolving.
If there’s a golden rule for responsible AI (and trusted technology in general), it’s this: It’s better to implement trust by design and ethics by design from the start rather than racing to close gaps after systems are up and running. That’s why, whatever your company’s maturity level with AI, including generative AI, it’s wise to put responsible AI front and center as soon as you can — and keep it top of mind every step of the way.
Helping you harness AI you can trust through frameworks, templates and code-based assets.
Lead with trust to drive sustained outcomes and transform the future of your business.
Generative AI is here and already transforming business. Contact us to learn more about this rapidly evolving technology — and how you can begin putting it to work in a responsible way.