Responsible AI isn’t just feel-good thing, nor is it just a litigation-avoidance strategy. it’s a growth strategy. It delivers tangible outcomes.
That’s the word from a recent PwC survey in which 310 business leaders gave their thumbs-up to responsible AI within their AI efforts. majority, 58%, cite improved return on AI investment, along with 58% crediting responsible AI for enhanced customer experience. At least 55% say it helps with innovation, and a similar number seeing enhanced cybersecurity and data protection.
Scaling responsible AI is the challenge. Half of executives, 50%, cite difficulty translating principles into scaled and operational processes, and a like amount are faced wtith cultural resistance to change. Thirty-eight percent are dealing with limited budgets or resources.
About six in ten respondents (61%) to a recent PwC survey say responsible AI is actively integrated into core operations and decision-making.
There is agreement across the industry that responsible AI needs to be part of every AI initiative. “AI is a business issue – not just an executive talking point,” said Cindi Howson, chief data and AI strategy officer at ThoughtSpot. “We all have a stake in this revolutionary technology and a shared moral and ethical liability to ensure AI isn’t simply a cool technology but also a technology that betters humanity.” Responsible AI, she added, “will take a village – deep collaboration that transcends the limitations of traditional policy-driven approaches.”
Responsible AI begins with employees at all levels. “It’s vital for employees to have clear expectations and guardrails to guide their AI usage to manage the risk involved,” said Danielle McMahan, chief people officer for Wiley. “Gather a group of internal subject matter experts to drive strategy and develop standards for ethical and responsible AI use,” she added.
Start by training employees to actually use AI, McMahan said. “Getting managers trained and on board should come first, as employees often turn to their direct supervisors for help.”
“The conversation can’t just be about capability,” said Jeremy Ung, chief technology officer at BlackLine. “It must be about trust which is the primary obstacle to AI agent implementation.”
For example, “in high-stakes environments like finance, where audit trails and accuracy are non-negotiable, agentic AI needs to be built on a foundation of verifiable, secure, and explainable systems, “ Ung pointed out. ”It’s the infrastructure, often overlooked but essential, that makes responsible AI possible: it is about clean data pipelines, robust APIs, and immutable logs.”
The next phase of responsible AI maturity “embraces a continuous innovation mindset—using technology to strengthen oversight while driving progress and performance,” the PwC authors predict.
“Don’t treat governance as an afterthought,” said Ramprakash Ramamoorthy, director of AI research at ManageEngine, a division of Zoho Corporation. “It begins with high-quality, unbiased data, explainable models, and auditable workflows. Equally important is establishing a human-in-the-loop review for high-impact decisions and continuous drift monitoring once models are deployed.”
Such AI ethics committees “should not be symbolic but operational, with clear escalation paths when a model’s decision deviates from expected behavior. Responsible AI cannot be delegated to one team; it is a culture that must be codified into every product and process that interacts with intelligence.”
