Enterprise hits and misses - NVIDIA makes a trillion dollar proclamation at GTC, agentic commerce isn't ready (yet), and robots are scary

Lead story - Four AI stories, four different AI lessons - this week, I took four very different lessons from the our latest AI coverage: 

Retail's future is tied up in agentic commerce, yes? Well, up to a point, says BestBuy, Wayfair, and Electrolux with welcome notes of caution struck - Stuart reviews the state of agentic commerce. My take: agentic commerce will definitely become an important purchase channel over time, but the obstacles to getting there are going to burn some early adopters. 

Robot futures – why robophobia is a bigger problem than technologists realise -  Chris examines another obstacle in robotic advancement: "The United Robotics research suggests that the more human our machine counterparts appear, the less people will accept them. This may be a challenge in the years ahead as our need for robot workers grows." My take: robotics is in an experimentation phase is most sectors, with adoption growth in narrow verticals. Experimentation is a healthy attribute, the hype factor is not. 

From ‘buy’ to ‘build’ – how FranklinCovey CIO Blaine Carter is replacing packaged apps with low-code and AI -   My take: Now more than ever, 'build' is a viable option alongside 'buy.' But as Ian notes in this low-code use case, you're not building from scratch; vendor platforms still factor in - as does risk management: "Carter agrees that the risks of switching from buying to building are real. But he suggests it’s not just a technical challenge but an operating model shift — one that changes how the IT team builds solutions and engages with the wider organization."

Tech disruption as a day job - Lloyds Banking Group CEO Charles Nunn on AI, digital transformation, and why his bank just isn't playing the same game as upstart fintechs. My take: as Stuart reports, Llloyds Bank now has bottom line benefits from its AI investments, as well as promising agentic cases underway. But Lloyds Bank is also dealing with the PR fallout from reports of an app security issue - a telling example that the downstream impacts of tech investments are always part of the equation. Moreso with AI... (for more on that, read on)

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

• Huangs for the memory – NVIDIA’s GTC tells us that the future has claws!  - It's early in event season, but GTC is one of the bellweather events. So what did we learn? After sitting through 18 product announcements and a $1 trillion chip revenue projection in two years, Chris had a few things to say: "Welcome to the future, folks: one of claws, relentless token manufacture (with no mention of the energy and carbon cost), and robot-run business. A future that generates so much wealth that a man who wears a $1.5 million watch while effecting not to has to pitch random trillions for a room full of Wall Street money men. Feeling confident? And the share price fell..."
• AI answer engines are changing how software gets discovered. Here’s how G2 is adapting - Barb looks at how G2 is adapting to buyer behavior shifts: "LLMs need trusted sources to cite in buyer searches. Without those trusted sources, buyers would have no confidence in the answers they get and stop using LLMs. G2 is evolving to ensure it continues to be one of those trusted sources."
• Workday rolls out Sana, its conversational AI gateway to enterprise work - Last week, Workday issued one of its most significant product announcements in years. Phil took in the virtual briefings, and issued his review: "Bhusri remains adamant that enterprises still value the deterministic certainty of Workday’s underlying system of record, but recognizes the importance of teaming this with the probabilistic reasoning that AI can layer on top." Also see: Phil's companion piece, 'We have to come up with a plan for them' - Workday leaders on the fate of workers displaced by AI. 

diginomica hits the road, as spring event season revs up: 

UiPath Fusion 2026 - Alyx assessed the UiPath agentic narrative in London with UiPath Fusion 2026 - if your data isn't ready, your agents aren't either: "What Ashley articulates so clearly is that the problem isn't the technology – the ceiling on value is the same thing it's always been: whether organizations actually understand their own processes and data well enough to give those systems something useful to work with." (More diginomica research on this topic to follow!). Also see: Alyx's use case: UiPath Fusion 2026 - how Travelodge is scaling intelligent automation. It starts with the business owner, not the bot. 

Qualtrics X4 - Derek pressed the issue on agentic AI via the Qualtrics X4 user event in Seattle, via Qualtrics X4 - the harder questions about agentic AI and experience context: "Qualtrics' central argument - that experience context is the missing ingredient for effective agentic AI - only holds if that context can actually flow into the systems where decisions get made. This is another challenging area for Qualtrics. Understanding that a customer is frustrated is useful. Understanding that frustration inside a ServiceNow ticket workflow, a Salesforce interaction, or a healthcare navigation agent is what the "system of decision" ambition actually requires." Also see: Qualtrics X4 - new CEO Jason Maynard declares the insight gap closed, and sets sights on driving outcomes.

IamPhenom 2026 - Phenom’s IAMPHENOM 2026 - quick event wrap - Brian crashed Phenom's event and came back with an AI-meets-vertical-recruiting review: "Finally, I liked Phenom’s approach to documenting and achieving value, much of it due to process improvements. That’s the right direction and I’d recommend Phenom develop even more process workflows and metrics that push customers into more advanced and aggressive work designs."

A few more vendor picks, without the quotables:

• Making AI work everywhere in the enterprise is the challenge - Accenture CEO Julie Sweet points to another quarter of booming Advanced AI demand - Stuart
• Blue Yonder expands its agentic AI footprint - from warehouse floors to Microsoft Teams - Derek
• GTC 2026 - Everpure tackles data readiness and flexible consumption for enterprise AI - Derek
• How Canva's design-specific foundation model unlocks editing of any AI-generated image - Phil

Jon's grab bag - Stuart debuts a new Friday diginomica feature: The long and the short of IT - the week in digibytes. Mark Samuels has a fresh customer use case with Snowflake: How Toyota Motor Europe’s integrated approach to data provides the road to solid AI foundations. Chris parses a problematic report: the UK's long awaited copyright missive (AI and copyright – UK government finally backs off 'free for all', but sits on its hands and offers no solutions). 

Cath explores the state of AI voice tech in Speaking of AI-generated voice technology, how do organizations get the tone right? I find the conclusion fascinating: "The overall finding of the Vocal Image report seems to be that users are happy to interact with AI-generated voices as long as they think they’re human." I won't question the data because I'm often an outlier, but I see it differently. I care much more about whether the voice AI understands me - and fulfills my request - than whether it sounds human or not. I downgraded from Alexa+ precisely because I found the Alexa+ bot too human-like, in a super-creepy, soulless way, and it didn't grasp my deterministic needs. Back to the old robot Alexa, that faithfully executes my carefully-crafted routines to perfection, and provides small comforts from a limited bank of predictable robotic responses ("I hope you had a great day!"). (Mis)adventures with voice AI are just beginning...

Best of the enterprise web

My top seven

• Meta's rogue AI agent passed every identity check — four gaps in enterprise IAM explain why - The idea that AI opens up new threat vectors is almost a cliche. Maybe a real-life wake-up call would spark urgency? Louis Columbus has one: "A rogue AI agent at Meta exposed sensitive internal data despite passing every identity check. Here are the four post-authentication gaps in enterprise IAM that made it possible." Also see: Columbus' OpenClaw can bypass your EDR, DLP and IAM without triggering a single alert.
• Constellation Research's Futures Forum: What CEOs are thinking - Constellation's Larry Dignan lays out the latest CEO findings: "
  "Companies need to identify and empower "responsible heretics," or people willing to challenge legacy systems, sunk costs and sacred cows."
• Why I’m Not Worried About Running Out of Work in the Age of AI - Dave Kellogg has a different angle on AI's job threat: "The one thing I find most missing from the AI future analyses I’ve read is a simple realization: in my experience, there is always, always more work yet to do."
• Agentic Commerce: A Genuine Paradigm Shift or Just Another Vendor Pitch? - Thomas Wieberneit wades into the agentic commerce hype: "There is a dangerous misconception that AI agents will completely replace human customer service teams. This is a recipe for disaster. No matter how advanced the model becomes, it will inevitably encounter an edge-case query it cannot resolve."
• When the Dashboard Says Green, but Your Program Is Drifting - UpperEdge's Ted Rogers on an underrated SI topic: project intervention. "Enterprise technology programs rarely fail in dramatic, obvious ways. They drift quietly, gradually, and often invisibly to the people who have the most at stake."
• Rethinking intelligence - Jeff Jarvis riffs on a notable new report on world models versus LLM limitations, co-authored by Yann LeCun.
• The Uncomfortable Truth About Enterprise AI in 2026: It's Not Intelligence, and That's a Problem - Thomas Wieberneit of CRM Konvos boils down my latest video debate with Dr. Michael Wu. I did my best to ruffle Dr.Wu's feathers on issues like grounding and faux intelligence, but Dr. Wu was ready for me. Wieberneit managed to stir the pot...

Whiffs

Meta's Metaverse isn't dead yet, but I wouldn't be buying virtual real estate in Horizon Worlds these days, not even swampland:

Okay, this is a bit awkward: 

(OpenAI can automate your business, but not theirs...). This is awkward too: 

And so is this...

See you next time.. 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.