PwC has published a new research paper in collaboration with Microsoft, exploring how businesses can more successfully implement AI solutions.

The paper, How to Deploy AI at Scale, outlines the steps and considerations that organisations should address before integrating AI and genAI into everyday operations, and highlights the importance of having a clear AI roadmap and strategy, as well as the critical role of cloud and cybersecurity in adopting responsible AI practices.

According to the playbook, firms should first look to set clearly defined AI goals that determine an overarching strategy and roadmap, aligning those ambitions with their AI maturity. The report claims that without this initial step, organisations fail to match business strategy with technical capability, leaving them without the robust architecture needed to build on effectively.

That leads to PwC and Microsoft’s next recommendation, which is to assess cloud infrastructures and consider how current cloud architecture meets future organisational needs.

According to an earlier study by PwC, only a small number of businesses have so far benefited from integrating AI and cloud technology, allowing them to more readily unlock new value from data as they build new AI capabilities.

However, the cloud infrastructure required to leverage AI’s capabilities presents new cybersecurity challenges, and according to PwC’s 2023 Global Risk Survey, organisations now rank cyber risk as the second most critical issue they face, just behind inflation.

That leads to the next steps in the playbook, securing the data needed to power AI, as well as streamlining cyber defences, including recommendations to utilise genAI in cybersecurity measures to assist with threat detection, response, and intelligence.

The paper also recommends that businesses ensure they keep up with the fast changing regulatory environment around AI, for instance the new safety and ethics frameworks recently introduced by the EU, while ensuring those responsible for AI rollout are on the same page.

PwC pointed out that a recent survey showed C-suite executives in different positions have opposing views on AI readiness, with 67% of CEOs reporting high confidence in their organisation’s ability to be in compliance with AI regulations, compared to just 54% of CISO/CSOs. 

One solution the playbook touts is to operationalise responsible AI practices, with businesses being encouraged to implement governance frameworks that monitor AI use in real time, ensuring continuous oversight and immediate response. 

Finally, the paper emphasises the importance of providing more AI training for employees, while allowing room for some AI experimentation among the workforce. According to the research, these methods build employee confidence in using AI tools, increase its adoption across the business, improve productivity, and help employers retain talent.

Recommended reading

• New Report Reveals Gen AI’s Potential Impact on Future Management
• IT Pros Divided Over How AI Can Improve Workplace Experience
• Ready or Not, AI Is Already In Your Workplace

“When implementing AI and generative AI technologies, it is essential to evaluate how they can strengthen your competitive edge and align with your overall business strategy,” said Anton Tseshnatii, a risk assurance lead at PwC Ukraine. 

“Equally vital is the continuous monitoring and adaptation of your AI systems to meet evolving business and security challenges. Ensuring widespread acceptance and successful implementation requires building trust in AI within your organization.”