What's Real with (Agentic) AI? And Who's Deriving Value?
Four Insights By the World's Largest Brands (Shared At IBM Think 2026)
Last week, I attended IBM’s annual flagship event, Think, in Boston alongside more than 6,000 senior leaders. For this article, I selected four examples that provide context for the question so many of you are asking: “Who’s doing what with AI? And what’s real?” At the end of each section, I share my personal take on the respective topic.
You can catch up on all the news and announcements in the official press release.
Disclosure: This post is sponsored by IBM. All opinions are my own.
Aramco’s AI Transformation Journey with IBM
Arvind Krishna, CEO of IBM, opened the first day by shining a spotlight on customers who are embracing AI systematically, far beyond a few isolated use cases. Sami Al-Ajmi, SVP of Digital and Information Technology at Aramco, shared one of these examples. Aramco and IBM have a long relationship dating back to the 1940s. Nowadays, AI and Quantum Computing are the top areas of collaboration between both companies
Ajmi got straight to the point and shared: “We are not interested in POCs and pilots. We want to move to production.” What stands out, though, is how he described his company’s approach: data, people, and infrastructure. Although no specific numbers were mentioned, from AI-driven exploration (upstream) to asset monitoring (downstream), cost savings and margin increases are the primary goals of AI use (and achieving them).
Aramco processes 10 billion data points every day, which is the foundation to build its own AI models. But it is the deep industry and domain knowledge that Aramco’s employees, of which 60,000+ are now trained in AI, leverage to derive business insights. Lastly, Ajmi shared: “[…] always, always tie it to business value.” (I couldn’t agree more with him on that one.)
My take: It’s important to hear how the world’s largest oil company is approaching AI strategically. Especially the emphasis on domain expertise, coupled with data and technology as a competitive advantage, is one that’s frequently missed in the AI debate in the U.S.
A Development Partner for the Enterprise: IBM Bob
The CEO keynote featured several announcements and launches, including IBM Bob, an AI-driven development partner for enterprises. Bob supports developers throughout the software development lifecycle, from analysis and planning to coding, testing, and documentation. Whether teams are looking to modernize existing applications (e.g., from one Java version to another), build new apps, or develop AI agents, Bob acts as a junior developer for a senior engineer and vice versa.
After the keynote, the show floor was swarming with attendees eager to see the newly launched product up close. I had a few more questions about IBM Bob and its differentiation in the enterprise, and sat down with Neel Sundaresan, GM of Automation and AI, the leader behind IBM Bob and watsonx Orchestrate.
Neel pointed to key features that IBM has added to meet the specific needs of enterprise teams. Whether it is a more robust set of questions before executing a command, analyzing one’s organization’s usage and spend, or the ability to pool and assign credits by team, enterprises can expect a level of transparency and robustness that many are looking for. With 80,000 developers already using Bob regularly, IBM has incorporated its learnings into the product. Bob is model-agnostic and selects the optimal model for the task behind the scenes, so developers can focus on building software rather than perfecting context management for LLMs or running limited model benchmarks.
IBM Bob also helps developers build AI agents according to the IBM watsonx Orchestrate agent development kit (ADK). IBM watsonx Orchestrate lets IT teams define governance and data controls, and its control plane simulates agent conversations, runs tests, and debugs agent scenarios for evaluations.
Learn more about IBM Bob in this press release.
My take: Positioning Bob as a development partner rather than an ‘AI agent,’ ‘agentic coding assistant,‘ or any other replacement for human labor resonates well with the developer audience. I took it for a spin, and Bob asked clarifying questions and asked for confirmation before performing any critical operations. I’m looking forward to seeing even more case studies and proof points in the future that highlight Bob’s unique market position.
The C-Suite AI Collaboration That Saved IBM $4.5B
Next up was an exclusive panel for about 200 senior leaders about how IBM achieved significant business benefits with AI by prioritizing “teamship” over titles across the C-Suite. It’s always easy to give advice to clients if you haven’t made the learning yourself. Because this panel had actually been through this change, I was especially excited to moderate the discussion and hear about the unique angles that three of IBM’s top executives shared:
Neil Dhar, SVP Americas Consulting at IBM, highlighted how CEO Arvind Krishna set the vision agenda, driving alignment across the senior executive team. The next task was to examine the workflows with the greatest opportunity to be transformed with AI. It wasn’t just one department or line of business that touched these, which drove collaboration with the CFO, COO, CHRO, and CIO offices. Together, they addressed strategic investments, operational changes, and the talent and culture blueprint to deliver the ROI.
Joanne Wright, SVP of Transformation & Operations at IBM, pointed out that AI accelerates whatever system you already have. Leadership must fix the system before scaling the technology. (It sounds obvious, but it is anything but…) IBM redesigned work around end-to-end value flows instead of functions, eliminating non-value-adding steps first, simplifying ownership next, and only then embedding AI for value creation. Joanne succinctly said that leaders need to shift from governing projects to governing outcomes with clear decision rights, prioritization mechanisms, and accountability.
Nickle LaMoreaux, SVP and Chief Human Resources Officer at IBM, emphasized that technology is just one aspect of value creation. It takes a cultural shift, and leaders must be empowered to embrace the unknown, even if it means facing the risk of failure.
My take: More often than not, CEOs and their boards still approach AI as a technology topic that the CIO or CTO team is asked to lead. Centralistic approaches often lead to siloed thinking, a lack of buy-in across the company, and the classic “not invented here” syndrome. In times of rapid innovation, change, and disruption, it takes the entire leadership team to succeed. It was good to hear IBM's approach.
Certifying AI Agents’ Skills with Digital Badges on Credly
In my second book, The HUMAN Agentic AI Edge, I talk about applying the Dual-Lens Principle. The idea is that we can borrow from tried-and-tested HR concepts when managing and governing AI agents. Updating an agent’s knowledge and verifying its correct recitation of information are among the dimensions. But whether an agent actually has the skills it (or its developer) claims remains unknown.
Dave Treat, CTO of Pearson, the company behind the digital credentialing platform Credly, shared how IBM and Pearson are collaborating to bring skill verification and digital credentials to AI agents. (Let that sink in for a minute.) The goal is for human team members to assess an agent’s skills before inviting the agent to the team: This process starts with gathering requirements and constraints, continues with documenting the current architecture, and closes with defining evaluation criteria. Organizations can choose the verification and autonomy levels. A dynamic assessment of the agent’s base layer (including behavior), skills, job fit, domain application, and performance evaluation is next before issuing a digital credential.
My take: Digital credentials elevate agents to first-class team members in a Human-Agentic-AI-enabled workplace. Official, documented credentials pave the way for internal and external marketplaces of agents with verified capabilities. Credentialing is one aspect of the emerging field of Agentic AI governance in the enterprise.
Conclusion
IBM showed several customers on stage that it was to pick just handful of stories to mention in this article. From Andre Agassi sharing how Agassi Sports brings world-class tennis coaching to anyone with the help of AI to Susan Doniz, The Walt Disney Company’s CIDO, talking about empahy in technology-driven transformations, and Carol McDaniel, Providence Health & Services’ VP of Talent Acquistion, highlighting the key learnings of the company’s Agentic AI journey, there are so many more stories to tell. You can watch the keynote replays here.
At the end, it’s Andre Agassi’s observation that stuck with me: “If you start this game wrong, you’ll never be able to become really good as you progress.” (I’m still thinking about whether he was talking about Tennis or AI…it certainly applies to both.)
Become an AI Leader
Join my bi-weekly podcast for leaders. Each episode features a different guest who shares their AI journey and actionable insights. Learn from your peers how you can lead artificial intelligence, agentic AI, generative AI & automation in business with confidence.
Join us live
May 12 - Reid Blackman (CEO of Virtue Consultants) is going to address the growing ethics gap in Agentic AI systems.
May 26 - Kris Saling (Chief Technology Advisor & Senior Data Leader) will discuss the top criteria when building multi-agent systems in HR.
June 09 - Joseph X Ng (Chief Strategy Officer of GeneGenius) will join to discuss emerging technology trends beyond AI.
June 23 - Doug Shannon (Senior AI & Automation Leader) will be on the show to discuss how AI is changing the speed of business. [More details to follow on my LinkedIn profile…]
Upcoming events
Join me or say hello at these sessions and appearances over the coming weeks:
May 11 - Guest on The Peggy Smedly Show to discuss the latest Agentic AI enterprise trends.
May 12 - Private event for members of a renowned global expert network.
May 14 - Lightning session at West Chester University of PA’s RECAP conference.
May 14 - Private event for hedge fund and asset management leaders in New York, NY.
May 26-28 - Attending the U.S. AI Congress in Washington, DC.
June 01 - Hosting an episode of O’Reilly’s What’s New in AI live show, covering the top AI news of the week.
June 29 - Hosting an episode of O’Reilly’s What’s New in AI live show, covering the top AI news of the week.
November 10-11 - Technology Sourcing in Chicago, IL.
Follow me on LinkedIn for daily posts about how you can lead AI in business with confidence. Activate notifications (🔔) and never miss an update.
Together, let’s turn hype into outcome. 👍🏻
—Andreas