Dimon and Jassy Set Out Their AI Spending as Surveys Show CEOs Now Lead AI Decisions

Title: Dimon and Jassy Set Out Their AI Spending as Surveys Show CEOs Now Lead AI Decisions

Date: 10 April 2026

Type: Blog

Author: SAASiQ (contact@saasiq.ai)

Word count: 1308 words

Reading time: 5 min

Published: 10-04-2026

Tags: #AI #Leadership #CEO #Strategy #Management #GenAI

Two of the best-known chief executives in US business used their annual shareholder letters this week to explain what they are spending on AI and why. Jamie Dimon's letter for JPMorganChase, published on 6 April, said AI will affect virtually every function at the bank and will definitely eliminate some jobs. Andy Jassy's letter for Amazon, published on 9 April, defended about $200 billion of capital spending this year, most of it on AI infrastructure. Both follow survey work published in January which found that at most large companies the chief executive, not the CIO, is now the main decision maker on AI.

CEOs now make the AI decisions

Boston Consulting Group's AI Radar 2026, published on 15 January, surveyed 2,360 executives across 16 markets and nine industries, 640 of them CEOs. Seventy-two per cent of the CEOs said they were their organisation's main decision maker on AI, twice the share of a year earlier.

Companies in the survey expect to double their AI spending in 2026, from 0.8 per cent of revenue to about 1.7 per cent. Financial institutions put their figure at 2.0 per cent and technology companies at 2.1 per cent, while industrial and real estate firms were at 0.8 per cent. More than 30 per cent of the 2026 AI budget is going on agents, the software that carries out a multi-step task rather than answering a single prompt.

Half of the CEOs said they believe their jobs depend on getting AI right. Ninety-four per cent of companies said they would keep investing even if AI does not pay back this year, and 90 per cent expect agents to produce measurable returns in 2026. BCG sorts respondents into groups, and the ones it calls Trailblazers put 60 per cent of their AI budget into upskilling staff, against 27 per cent for the Pragmatists and 24 per cent for the Followers. Trailblazer CEOs also spend more than eight hours a week learning AI themselves.

What the returns look like so far

PwC's 29th Global CEO Survey, published at Davos on 19 January, asked 4,454 CEOs in 95 countries and territories what AI had done for their companies over the previous 12 months. Thirty per cent reported higher revenue from AI and 26 per cent lower costs, while 22 per cent said AI had raised their costs. Fifty-six per cent had seen neither a revenue nor a cost benefit. Twelve per cent had seen both.

That 12 per cent were two to three times more likely than the rest to have embedded AI extensively in products and services, demand generation and strategic decision-making. CEOs whose companies had what PwC calls strong foundations, such as a responsible AI framework and a technology estate that lets AI be used across the whole business, were three times more likely to report meaningful financial returns. Separately, only 30 per cent of CEOs were strongly confident about revenue growth over the next year, down from 38 per cent.

Deloitte's State of AI in the Enterprise report, published on 21 January, drew on 3,235 business and IT leaders in 24 countries, surveyed in August and September 2025. The share of workers with access to sanctioned AI tools had gone from under 40 per cent to about 60 per cent in a year. Only 25 per cent of organisations had moved 40 per cent or more of their AI pilots into production, although 54 per cent expected to get there within three to six months. Nearly three quarters planned to deploy agentic AI within two years, and 21 per cent said they had mature governance in place for agents.

What Dimon wrote

Dimon's letter says the pace of AI adoption 'will likely be far faster than prior technological transformations, like electricity or the internet'. He wrote that AI will affect virtually every function, application and process at JPMorganChase, and that in the long run it will have a huge positive effect on productivity.

He wrote that AI will definitely eliminate some jobs while it enhances others, and that the firm will have definitive plans to support and redeploy the staff affected. He also argued that business and government can do a good deal to encourage retraining, income assistance, reskilling and early retirement for people whose work is displaced.

CNBC reported that Dimon does not regard AI investment as a speculative bubble, although he said the eventual winners and losers cannot yet be predicted. The letter lists the new risks AI brings, from deepfakes and misinformation to cybersecurity weaknesses, and warns that misuse of customer data is likely to get worse with 'agentic commerce', where AI agents buy on a customer's behalf. Yahoo Finance's report on the letter put JPMorganChase's planned technology spending for 2026 at $19.8 billion, covering AI, data infrastructure and cloud.

What Jassy wrote

Jassy's letter, posted on 9 April, gave a number Amazon had not published before: AWS's AI revenue run rate was over $15 billion in the first quarter of 2026. He compared it with AWS itself, which had a run rate of $58 million three years after it launched commercially.

'We're not investing approximately $200 billion in capex in 2026 on a hunch,' he wrote, pointing to OpenAI's commitment of more than $100 billion and to other customer agreements that are signed but unannounced, or still being negotiated. Amazon's chips business, covering Graviton, Trainium and Nitro, has an annual revenue run rate of more than $20 billion and is growing at triple-digit rates. Jassy expects Trainium to save tens of billions of dollars of capital spending a year once it carries inference at scale.

The letter also names the agents Amazon now sells to business customers: Kiro for coding, Transform for software migrations and Quick for knowledge work. Jassy wrote that as customers expand their use of AI, they want their inference to sit near their other applications and data.

What Oracle has put in Fusion

Oracle made two sets of Fusion announcements in the same fortnight. At its AI World Tour event in London on 24 March, Oracle announced 22 Fusion Agentic Applications across ERP, supply chain, HR and customer experience. They work inside Fusion's existing security model, with the same role-based access, approval hierarchies and permissions a user has, carry routine work forward and pass exceptions to a person. The finance example Oracle gave was the Collectors Workspace, aimed at cash collection and days sales outstanding.

On the same day Oracle added an Agent ROI dashboard to AI Agent Studio. It reports time saved, cost savings and productivity gains for each agent, broken down by workflow, team and function. The release also added support for models that handle images, audio and video, real-time monitoring and observability, workflow orchestration with rules-based controls and human oversight, and contextual memory, so agents keep context between interactions. Oracle says AI Agent Studio is available to Fusion Applications customers at no additional cost.

On 9 April, at the AI World Tour in New York, Oracle announced 12 more agentic applications, eight for supply chain and four for finance: Claims Settlement Workspace, Collectors Workspace, Sales Order Command Center and Cost Accounting Close Workspace, the last aimed at prioritising work and shortening the period close. Oracle said they are available now in Fusion Cloud ERP and SCM.

In the third quarter of Oracle's fiscal year, reported on 10 March, cloud applications revenue, which includes Fusion and NetSuite, was $4.0 billion, up 13 per cent, against 84 per cent growth in cloud infrastructure. Fusion Cloud ERP brought in $1.1 billion.

What happens next

SAASiQ's view is that a Fusion customer whose CEO now signs off AI spending should have the ROI dashboard running for each agent before it goes live, so that time and cost figures exist from the first day rather than being reconstructed later.

JPMorganChase reports its first-quarter results on 14 April.

SAASiQ - Intelligent Solutions for SaaS ©