Enterprise hits and misses - CIOs reckon with AI use cases, and bring on the agentic substrate debate

Lead story - CIO AI gut check - hot topics and use cases

Where do CIOs stand on AI? It's complicated... (See the diginomica network research - CIOs navigate AI's weight of expectation). 

This week, Mark Chillingworth put three CIO stories into the mix, all with a different AI angle. Start with AI Bots - a new risk and opportunity for CIOs to manage. The problem? As per Mark: 

CIOs and CTOs in our network are seeing an increase in automated traffic to their infrastructure and are worried. These digital leaders represent a diverse set of sectors such as property, financial and professional services, tech firms, academia, manufacturing, media, and charities.

This is not an inexpensive annoyance. One CIO reports a six-figure increase from unwanted traffic. Mark quotes Tom Howe,  Director of Field Engineering at Hydrolix:

What was discovered is that bots were crawling sites in a novel way and found a set of high-resolution images on a web server, and due to the non-deterministic behavior of the bots, it was then downloading the images. The reality is, even if bots are not malicious, they can lead to very unexpected consequences.

There isn't a cure-all answer, but observability is clearly at the center:

CIOs will need data from these tools turned into insight that will inform business and AI bot decision-making. Observability and bot insight tools are going to become an essential part of the IT estate.

Next up? Ground Control CIO Chris Howell re-thinks quoting to drive growth and cut carbon. Project after project, CIO themes seem to come back to transformation - or business relevance. As Mark writes:

[Howell] describes transformation as having two halves that CIOs must navigate. The first half is about doing transformation to people; this is the heavy lifting of implementing infrastructure and a new ERP, for example. The second phase then involves everyone coming together, he says, and delivering the business model and customer service benefits.

And how does AI factor in? For Ground Control, it's a practical, problem solving angle. Mark quotes Chris Howell:

Our job as CIOs is to put the information into the hands of our people, but the satisfaction and the payoff is solving a problem and enabling someone to use data, insight, or a new tool, and that is the dopamine hit for me. When we solve a problem.

One more for the mix: CIO Mark Bramwell and Said Business School’s educated adoption of AI. As Mark explains, Bramwell needed to make AI adoption economically viable. This mean negotiating a token and tool-based consumption model for AI use across the business school (despite the need for budget predicability) As Mark Bramwell explains:

Variable cost is the enemy of higher education, given the tensions on our budgets, but it was a conscious decision to move away from AI models that are per-user, per-month licensing. When you look at the range of AI tools available, we cannot provide a portfolio of choice when a licence can cost anything between $8 to $30 per-user, per-month.

These CIO AI lessons converge: AI adoption is almost always a priority, but so is business relevance, as in: delivering IT results. How companies navigate AI adoption will vary by culture/industry/region, but the tension of costs-versus-results will remains. 

At the risk of belaboring the obvious, solving pressing business problems will help a lot. Otherwise we are stuck in slide deck mode on the one hand, and generic AI use cases on the other (see: Copilot). 

Diginomica picks - my top stories on diginomica this week

• AI and grocery - US giant Albertsons gets personal in its digital intent - AI and... Grocery stores? For Albertsons, that means digital customer experience, merchandising intelligence, labor optimization and supply chain optimization. Stuart assesses the latest.
• Beyond the sea - transforming an extreme market is hard. Nauticus Robotics CEO dives into how it can be done - Chris forays into the frontier of robotics again, and emerges with another reality check: "This, then, is the recurring reality of the robotics sector: an exciting vision of a real-life Transformer and an Uber-like market – seamlessly changing itself and the market’s slow-moving economics – rapidly replaced with something that offers a simpler, well-engineered solution that works within existing structures."
• Vector search is table stakes - OpenSearch Foundation bets on data sovereignty and enterprise trust instead - Alyx looks to bridge that open source enthusiasm versus enterprise procurement gap: "What Lewis and Meadows described is not a search engine pitch. They are building conditions under which enterprises can own their AI data layer without vendor lock-in."

Vendor analysis, diginomica style. Here's my three top choices from our vendor coverage:

• Canva Create - Canva launches its mission to democratize AI automation in the workplace - Phil makes sense of Canva's enterprise ambitions - which are coming together with surprising speed.
• Can AI infrastructure costs be a value driver? Hot topics and customer views from Oracle AI World 2026 - Oracle's AI World pit stop in New York city gave me the chance to press questions in candid 1:1s. Here's my best-of, which includes details on the AI infrastructure cost angle, and (self)disrupting with partner agent/app building. Also see: Derek's latest use case: Wolseley's CTO on Oracle Fusion and AI - start with the easy 80%, and bring up the bottom third.
• Why is enterprise AI stuck? OpenSearchCon Europe 2026 says the bottleneck has moved from the model to the data - Via OpenSearchCon, Alyx bears down on the critical AI issues of 2026 - and they aren't about the model. 

Salesforce TDX 2026 coverage - headless is here (for real this time) - with Ian on the ground, and Stuart firing off his trademark (virtual) missives, we covered a headless TDX is a multi-headed way. But why headless now? As Ian explains, this is not a purely technical move: 

Instead I view this refactoring as evidence that Salesforce is serious about the shift it is evangelizing towards the agentic enterprise – and that it is basically showing its customers what that shift looks like in very real – and very concrete – terms. Not because it is adding agents, but because it is restructuring itself to be legible to them. And Headless 360 is the mechanism for doing that.

• Salesforce TDX 2026 – why Salesforce’s Headless 360 announcement at TDX is really about operating model transformation - Ian
• Salesforce TDX 2026 – how AgentExchange moves the point of purchase closer to actual work - Ian
• Salesforce TDX 2026 - headless thinking for the Agentic Enterprise era as developers face the shift from determinism to probabilism - Stuart

A few more vendor picks, without the quotables:

• Use of agentic AI erodes GDPR compliance as we know it. Wipro's 'privacy by design' comes into its own - Katy
• Certinia unveils Veda, its AI engine for Professional Services Automation - Phil
• A new Answer Engine Optimization tool, plus other updates - shooting the breeze with HubSpot's Spring 2026 Spotlight - Barb

Jon's grab bag - Chris has another (well-earned) scorcher on AI and copyright: AI and copyright – UK Government still refuses to engage with creators, claims artists rights leader. What will it take? Stuart has another batch of news we shouldn't have missed in The long and the short of IT - the week in digibytes. Alyx issues another EU Cyber Resilience Act wake up call: Cloudsmith warns - most teams won't meet the EU Cyber Resilience Act's software supply chain deadline. 

Madeline says what needs to be said in Something for the weekend - when it comes to women in tech, stop asking us about childcare and babies, and pay us properly! Meanwhile, Chris valiantly parses the physical AI hype again in Let's get physical (again) - Capgemini starts a heavyweight fight with a bantam-weight AI argument.

Finally, we've got another Executive Intelligence podcast for your perusal: Derek's on deck with Executive Intelligence podcast - Celonis President Carsten Thoma on why enterprise AI needs operational truth before it can deliver. The future of software remains on the (hot) front burner: "Thoma's take was that tools lacking depth, differentiated data or a critical mass of business context are at risk. But he also argued that the large-scale vendors with deep operational knowledge and history will evolve." Check the full convo...

Best of the enterprise web

My top seven

The Agent Wars Are Over. The Substrate Wars Just Started - so how will we assess enterprise software vendors? 

Thomas Wieberneit has an important post for us to ponder, breaking down the agentic data strategies of SAP, ServiceNow, and Salesforce - but you could really apply these dilemmas/opportunities to most big enterprise software vendors. Wieberneit writes: 

I look at all these titan moves through two lenses.

• Interface control: who owns how users and agents access business applications.
• Orchestration: who owns the layer that coordinates work across systems.

I like this frame, but I propose a customer-centric one as well:

Who provides the best end-to-end process/data layer? (Because agents will continue to have fragile context when they must communicate across vendors... It's a problem that can be mitigated, but end-to-end platforms have advantages, particularly in agentic workflow automation).

Who provides the most effective agentic orchestration platform, which (hopefully) standardizes agent behavior for customers managing multiple vendors, which is, basically all of them?

Who is the best at turning "messy" enterprise data silos into relevant/consumable context for agents, with the least amount of friction? 

Who provides the best context and semantic/ontology layers, including unstructured and external data? (The vendor that provides the best data layers for helping agents "understand" (ahem) the most relevant data will have a big edge, and will have strategic edge over incumbents who can't build/partner on this). Notice I did not say "grounding" - grounding LLM agents is an architectural oxymoron! At best, you hope to guide them...

Who provides the best agentic user experience? (Alas, a better user experience around agentic queries may prove more attractive that the most contextually-accurate, as we've seen with AI-inline search versus traditional search). 

I hear the objection: no one vendor can nail all of this down. Agreed - that's where the smartest operational decisions will have to be made. One end of the continuum: continued deep relevance as a go-to strategic vendor. At the other end is: becoming a data call for another vendor (legacy SaaS data repository alert!). Where you land will depend on many factors, but I'll put user experience, data trust and agile/adaptive/extendable platforms at the top.

Whoops, one more: who has AI pricing that my CFO can actually understand, with a smart balance between cost predictability and consumption? (Err, no one but...)

Oh, and if "substrate" becomes an enterprise buzzword, I'm calling it a career...

• Most enterprises can't stop stage-three AI agent threats - new agentic threat vectors are keeping Louis Columbus busy I'd say: "88% reported AI agent security incidents in the last twelve months. Only 21% have runtime visibility into what their agents are doing."
• Mythos and Cybersecurity - Bruce Schneier has a follow-on post on Anthropic Mythos, and the constant upping-of-the-security ante with more powerful models.
• AEO, GEO, and What Happens as AEM Enters the Chat - yes, we've got ourselves from jargon soup. Don't care - this post from Constellation's Liz Miller is the best content/marketing/search strategy post of the year.
• Let’s Form a Coalition to AI Smart – 'Supply Chain Shaman' Lora Cecere has some more zingers for us! "If we are not clear on outcomes, how can we be autonomous?"
• Why the “AI Job Apocalypse” Narrative is Wrong - I need to dig into the Anthropic jobs study for you a bit further, but for now, let's go with this one from November...
• Multi-media picks: so, Brian Sommer and I did a vigorous edition of our Enterprise Month in Review video show, with an audience ready to let it rip: Enterprise Month in Review - event detox and AI agent reality check edition.

Whiffs

So is this a whiffer, or a really savvy pivot? 

Then there is this super fun new venture: 

Finally, it sure seems like we are well into the annoyance economy, eh? 

If you find an #ensw piece that qualifies for hits and misses - in a good or bad way - let me know in the comments as Clive (almost) always does. Most Enterprise hits and misses articles are selected from my curated @jonerpnewsfeed.