Focus on AI: Putting It to Work for Your Organization

Many artificial intelligence initiatives have followed a path similar to the Wright brothers’ early efforts at flight, said Brent Blawat, an AI strategist with CDW. The first flight crashed after 3 seconds aloft, and the next one flew for all of 12 seconds. Similarly, Harvard Business Review reports that an estimated 80% of first-time AI deployments end in failure.

But to have success, organizations must keep trying. That seems to be what’s happening, as 63% of organizations planned to increase their AI spending in 2024. “If you’re not budgeting now for AI, you probably should be, because AI is not going away,” added Blawat, speaking recently at the CDW Executive SummIT on “Building Today’s Technology to Enable Future Success” in Phoenix.

AI was a topic on nearly every participant’s mind at the SummIT. Speakers made clear that organizations engaged in AI initiatives must approach the technology with a clear strategy that accounts for business objectives as well as security.

As organizations begin to pilot AI initiatives, they are looking to avoid the pitfalls that have scuttled the efforts of others. A strategic approach is necessary to maximize the chances of success.

IT leaders have numerous factors to consider as they develop their AI strategies. For example, AI capabilities can be delivered via a variety of models, including on-premises technology deployments, as well as cloud-based infrastructure, Platform or Software as a Service implementations.

CDW can help organizations find the model that best meets their needs, Blawat said, adding, “This year has been busy for us AI people in that we’ve learned a lot of lessons. The pace of change is astronomical.”

Richard Galvez, a principal field solution architect with Pure Storage, recommended that organizations start small with use cases that improve worker productivity. IT teams must work with other lines of business to establish frameworks for evaluating the success of AI initiatives and ensuring their safety.

To support their AI efforts, organizations should invest in modern infrastructure that can handle the processing and storage demands of the technology, Galvez advised, and they should scale up initiatives only when their value is clear.

As AI technologies have matured, the security implications have come into focus. Cybercriminals are making use of AI to increase the number and sophistication of their attacks, said Niall Browne, senior vice president and CISO for Palo Alto Networks. For example, there has been a 1,265% increase in the volume of phishing emails since ChatGPT launched in November 2022.

To address the risks, organizations should ensure that their AI initiatives meet their standards and policies. Organizations are also starting to develop policies for the responsible, ethical use of AI. Fortunately, Browne said, IT leaders can employ AI tools to help them address their security challenges, adding, “You don’t want to bring a knife to a gunfight.”

As organizations consider AI, they should ask themselves some important questions, Blawat said:

• Where does my inference data go?
• How is my inference data secured?
• Is the platform using secure infrastructure?
• What security controls are in place for the platform?

Addressing these questions early can help IT teams avoid problems down the line, he said.

Ultimately, the real value from AI will come from how organizations use it. Blawat highlighted retrieval automated generation — a technique that allows organizations to talk to their data using a chatbot — as a valuable use case. For example, for many organizations, responding to standard questions about benefits can take human resources employees hours or even a day, but RAG automates responses to these requests, enabling companies to return answers in a matter of seconds.

Finding the right use cases for AI is another aim of an organization’s strategy. “We’re taking a very measured approach with AI,” said Sanjay Sood, senior vice president and CTO with CDW. “We’ve invested in various technologies such as robotic process automation. We’ve been thinking about AI for a long time. We’re very methodical about where we want to place our bets.”

One key investment CDW has made has been in Microsoft’s Copilot AI tool. CDW has rolled out more than 8,000 licenses for Copilot to help users handle tasks such as summarizing meetings.

As they begin their journey with AI, many organizations don’t know where to start. CDW can help them take advantage of AI through several services. A Data Readiness Assessment helps organizations address issues related to data quality, privacy, stewardship and lifecycle management. An AI Enablement Workshop helps organizations identify the use cases that would be most helpful for them. “Not everything should be AI,” Blawat said. “We help organizations level-set so they can better understand what AI is and what it is not.”