LONDON: IBM launched on Tuesday at its annual Think conference in Boston what it described as its most comprehensive set of enterprise artificial-intelligence products to date, and also argued that there is a competitive chasm growing between companies that have integrated AI into their core operations and those still running pilots.
CEO Arvind Krishna spoke to media ahead of the four-day event, unveiling what IBM calls an “AI Operating Model” — a suite of tools built around four areas: how AI agents are managed, operations are automated, business data is made available in real time, and governance is maintained across complex regulatory environments.
Krishna said the focus this year is on helping enterprises move beyond isolated AI projects and into what he called a full operating model — one where AI drives productivity and revenue growth across the entire business.
“By 2030, the companies that are doing well will get 40 percent at the price productivity using AI. The more important part, 60 percent are going to get new sources of revenue, new sources of business models,” he said.
“But you have to act where the data is, and over 70 percent of all the data is still sitting inside the enterprise, in systems that are core and germane to them.”
The centerpiece of the agents push is an upgraded watsonx Orchestrate platform, which acts as a control layer coordinating AI agents built by different teams across an organization, ensuring they remain accountable even as their numbers scale.
IBM also launched IBM Bob, an AI development tool already deployed internally across tens of thousands of staff and is now available commercially. Over 80,000 IBM employees self-reported a 45 percent increase in productivity using the tool.
“We decided to release this to the market, and the uptake has been dramatic, largely through e-commerce actually so far,” said Rob Thomas, IBM’s senior vice president of software and chief commercial officer.
“Users and clients see the value of what I would call multimodel, orchestration and optimization. You get all the benefits of the best models without always having to pay the most money to get the performance that you need.”
Arvind Krishna, Chairman and CEO, IBM Corporation speaks during the 2025 Concordia Annual Summit in New York. (AFP/File)
The company does not compete to build frontier AI models, instead routing across a portfolio of partners — including Anthropic, OpenAI, Mistral, and Meta — while maintaining its own smaller Granite models for on-premises and regulated use cases.
On data, IBM showed the fruits of its $11 billion acquisition of Confluent, completed in March, which now enables AI systems to draw on live, continuously updated business data rather than the static snapshots that have hampered previous deployments.
A new data-processing engine built with NVIDIA cut costs by 83 percent and delivered performance 30 times better for Nestle across operations in 186 countries.
“AI is really about productivity, and your AI is only as good as your data, which informs everything that we’ve been doing across both AI and hybrid cloud,” said Thomas.
For businesses in regulated sectors, IBM announced the general availability of IBM Sovereign Core, which embeds compliance directly into infrastructure so it cannot be overridden as laws evolve. Launch partners include Dell, AMD, and French AI firm Mistral.
The announcements come as IBM reported a 15 percent rise in first-quarter net income to $1.2 billion, though slower revenue growth compared to the prior quarter had rattled investors ahead of Think.
On quantum, IBM closed with what may prove its most consequential announcement of the week.
Scientists from IBM, Cleveland Clinic, and Japan’s RIKEN institute have simulated a protein complex of 12,635 atoms using IBM quantum hardware — the largest such simulation of a biologically meaningful molecule on record, with direct implications for drug discovery.
Krishna, who has committed to delivering the world’s first large-scale fault-tolerant quantum computer by 2029, pointed to small-molecule pharmaceutical research as the most immediate commercial frontier.
Moderna is already working with IBM to explore quantum applications for mRNA medicine. The science, Krishna said, is no longer a lab experiment — it is a countdown.
“Quantum is now literally just around the corner,” said Krishna, though he was careful to temper expectations. “I would not point to quantum as a large commercial opportunity yet.”
