tag:dlvr.it,2018-03-23:ac6861e9e2d249fc92b2895b5ff7a2e0 2026-09-27T19:06:35-07:00 tag:dlvr.it,2026-09-27:cfba738c4e9256a7b83a4adb8b720636 2026-09-27T19:06:35-07:00
"Digital sovereignty describes how much control an organization has over its infrastructure, data, operations, and technology choices. Sovereignty is a spectrum, and architecture decisions can move an organization a step in either direction. Recent research by SUSE suggests that almost all enterprises are prioritizing digital sovereignty, but only 52% are actively taking steps toward it. That gap is largely an execution problem, and much of it surfaces in everyday platform decisions. If your team supports regulated industries or deploys in on-premises or air-gapped environments, you may be especially familiar with growing pressures around sovereignty." ]]> tag:dlvr.it,2026-09-27:3691fa650c01452791b0e8e6c9b561ed 2026-09-27T18:52:28-07:00
"Meta’s Muse is a mainstream version of AI agents already popular in Silicon Valley for automating personal tasks, like OpenClaw and Instinct. It’s a hit for the company: the Muse app has been downloaded over 900,000 times in its first week, according to Sensor Tower. The user experience is less like prompting a chatbot, and more like texting a friend with task requests—the AI agent acknowledges messages with a thumbs-up emoji and gets to work in the background. Muse is also available through Meta’s WhatsApp platform." ]]> tag:dlvr.it,2026-09-27:bbfefd0a480fdaee309db7a195cc4bb4 2026-09-27T18:38:58-07:00
Cloud ERP and AI disruptions - where are the real moats, and is a partner ecosystem one of them? Acumatica makes its case https://ift.tt/gRoUH8K This summer, Acumatica acquired Vertrax, an industry vertical partner. But the questions run deeper: in the AI ERP era, is industry functionality still differentiating? And how is Acumatica disrupting itself with its partner AI app building, while keeping up a clear roadmap for all? It's a great time to press the questions. via Jon Reed – diginomica https://diginomica.com September 24, 2026 at 02:30AM ]]> tag:dlvr.it,2026-09-27:118a812b6dc0ea3d36239683e14b048b 2026-09-27T18:25:14-07:00
"The incident also highlighted a separate enterprise problem: Muse can connect to services using credentials the user provides and act on the user’s behalf, while Muse itself gives enterprise security teams little central visibility into that access." ]]> tag:dlvr.it,2026-09-27:f1d08ba505c1372dad2dff38adcedf45 2026-09-27T18:25:13-07:00
"A guardrail tries to influence what a model chooses to do. A boundary determines what it can do when it chooses wrong. You cannot prompt-engineer away an exposed Docker socket or fine-tune a model out of a badly configured package repository. That however does not make alignment optional – I want to make it explicitly clear. Alignment reduces how often the boundary gets tested, while containment limits the damage when it does get pushed. But containment also has to work when the model itself is behaving as intended, because good behavior does not protect an agent from bad input. A poisoned web page, malicious issue comment or compromised documentation can steer an agent toward harmful action without the model becoming “misaligned” in any meaningful sense." ]]> tag:dlvr.it,2026-09-27:2649e6a4a65938e66a6d09666d56b9a8 2026-09-27T15:29:04-07:00
""Nothing works right if your CRM is out of date, so we fundamentally reimagined it for the AI era," the company wrote about the new capability. The real test will be just how accurate the AI is when it comes to entering data and how much sales teams can trust the information it's pulling. If it works as described, it will solve a major pain point, but if it makes a lot of mistakes, humans will be left to clean up after it, and not much efficiency will be gained." ]]> tag:dlvr.it,2026-09-27:cb29610178226795554ea62009152327 2026-09-27T15:29:04-07:00
" Industry Knowledge Graph: A structured representation of the knowledge, relationships, entities, and rules that define a particular industry. Think of it as a map of how an industry works, rather than simply a database of industry facts. Semantic Layer: The translation layer between raw enterprise data and the business meaning people—and increasingly AI—need to use the data. Ontology: Representation of a series of interconnected truths. Process Canonical: Workflows between parties. A process canonical is a standardized reference model for how a business process is defined and structured, providing a common language across systems, functions, and organizations. Operating Model: An operating model defines how an organization is designed to operate while a process canonical defines how a particular business process is standardized and represented. Unified Data Model: A unified data model describes how data is structured and related. While an operating model describes how the business is organized and works. For example, Distribution Requirements Planning (DRP) and Transportation Management Systems (TMS) do not have a unified data model. " ]]> tag:dlvr.it,2026-09-27:4ed9913840277895f3d54cb65062d117 2026-09-27T15:29:04-07:00
"Once you see the problem that way, it becomes clear why the individual negotiations cannot be optimized independently. A concession can look excellent inside one agreement and still make the transformation worse off overall. You can negotiate very attractive cloud economics while creating an architectural dependency that limits your model choices. You can secure aggressive SI rates while allowing the SI to establish a proprietary delivery structure that makes another integrator impractical later. You can win a contractual right to terminate and then discover that the data, configuration, operational knowledge or orchestration required to exercise that right sits somewhere else in the provider network." ]]> tag:dlvr.it,2026-09-27:e42efcceffca1c5518e5764dabd2c964 2026-09-27T15:02:10-07:00
"Cobalt CISO Andrew Obadiaru described the problem in written answers to VentureBeat. “What concerns me is that permissions without strong identity eventually create an accountability problem,” he wrote. “If several agents are operating through a shared service account, API key or user-delegated credential, you may be able to restrict what that credential can do, but it becomes much harder to answer basic security questions. Which agent actually took the action? Who authorized it? Can I revoke that agent without disrupting everything else using the same identity?” ]]> tag:dlvr.it,2026-09-27:0acbd6234a43d1702e2283463d699544 2026-09-27T15:02:10-07:00
"When Ralf asked which signal carries the greatest predictive value, Andreas gave the best answer of the evening: the biggest threat is the thing you do not hear. Satisfied customers go quiet; so do customers about to leave, because they do not want you on alert. His most overrated customer signal? Kindness. Inexperienced sellers mistake it for a buying signal. And it is not. This is where the AI story has a hard ceiling. Sentiment analysis can pick up words that people typically use when they are annoyed, even when they say them politely, but in Europe it requires consent and in any case it only works on conversations that happen. Silence produces no transcript. A buyer who has been thinking about a new CRM for five years produces no data in your system until the day they call. As Andreas put it, that is the first day for you, not for them. AI "knows what it knows," and it does not know what happened inside the buyer's organization before you arrived." ]]> tag:dlvr.it,2026-09-27:5caae0189f6dbabc7ceb57e0f5c5b5d7 2026-09-27T15:02:09-07:00
Enterprise hits and misses - AI needs more than ROI - it needs buyer trust. AI leadership needs a rethink, and Dreamforce is a wrap https://ift.tt/B8bal7H Readers pressed me on buyer trust in AI - time to rethink. ROI is the higher bar for AI, but is it the best metric? AI alignment, tokenomics, and doomerism... An interesting backdrop to Dreamforce, and our coverage roundup is here. Your whiffs include a new (strange) buzzword: AI (work)slop grenades. via Jon Reed – diginomica https://diginomica.com September 21, 2026 at 06:53AM ]]> tag:dlvr.it,2026-09-20:3059918c236fcdf67a1fd6dbabc7d117 2026-09-20T21:15:47-07:00
"ROI is therefore a poor primary measure of transformational business value when the objective extends beyond improving an existing activity. Legitimate returns from automation, lower costs, or faster processes show that an activity improved, but they do not establish whether the enterprise created a better operating model, changed how decisions are made, developed new sources of growth, or built capabilities competitors will find difficult to replicate. Business Value must capture observable improvement in how the enterprise grows, operates, decides, serves customers, manages risk, and adapts, including financial results, without reducing value to them." ]]> tag:dlvr.it,2026-09-20:3e3c08743e08a9c8a8d89492df0a1b01 2026-09-20T21:15:47-07:00
"Even the best-performing coding agents fail more than 60% of the time on tasks drawn from real codebases. Knowing the agent failed, though, is different from knowing why. “It’s not enough to just look at the logs or the inputs and the outputs,” el Hallak tells The New Stack. “It is important to figure out how it got to the answer. What were the reasoning traces? What tools did it utilize? Where did it get stuck? Where did it decide to try a new approach?” That can require replaying the agent’s execution to see where it went off course. What looks like a model failure may have started somewhere else in the stack. And that’s the tricky part for developers. Agent bugs aren’t always model bugs." ]]> tag:dlvr.it,2026-09-20:5b99b37592c59a4c3a9b544932eafc70 2026-09-20T21:02:45-07:00
"Kayne McGladrey, senior IEEE member, independent vCISO, and author of "Cyber Risk is a Myth," told VentureBeat that the privilege problem predates agents by decades. “We’ve had this pattern for 40-plus years,” McGladrey said. “Take this user account and clone it to that user account. We’re now doing that with agentic systems, except it’s an agent, not a human, and as we’ve seen with agentic AI, it does whatever it needs to do to get its job done, and sometimes that uses far more permissions than it should.” PaperCut’s service accounts fit that exact 40-year-old pattern. Domain-level privileges, no expiration date, no identity governance wrapper. The agents found them and used them." ]]> tag:dlvr.it,2026-09-20:f35bea619e0b0c60164344d4fbc1de4a 2026-09-20T21:02:45-07:00
"decisions, taking actions, and automating decisions. The kill switch should isolate the system and cut the AI off from other systems, the outside world, and data. Security access should be revoked and security tokens and API keys should be blocked. These safeguards should operate independent of the AI's own reasoning software, operate in physical and digital forms, and should not be overridden by AI. The recently introduced US Ai Kill Switch Act, if enacted, would provide a significant foundation for AI safety." ]]> tag:dlvr.it,2026-09-20:d9f0323b87e70cbd021337130d1ba5f9 2026-09-20T16:04:25-07:00
"Reckitt Senior Vice President Global Business Services & AI Bastien Parizot joins the diginomica network podcast to explain why the consumer goods giant behind Dettol, Durex, Finish and Nurofen is using its shared services division as the driving force for AI adoption. Parizot discusses the proof of value test that replaced proof of concept at Reckitt, the rollout of AI to thousands of marketers and R&D practitioners across marketing, R&D and procurement, and how digital simulation has gone from a few hundred runs a year to tens of thousands - making packaging more sustainable in the process. He also reflects on why he won't wait for perfect data before deploying, the human in the lead principle that governs Reckitt's AI and hallucination risk, and whether AI can shift the dial in a tight-margin industry." ]]> tag:dlvr.it,2026-09-20:1deb86e9d43c5ee3d714d24abc44cdb3 2026-09-20T16:04:25-07:00
"AIforce also does not undermine the seat. A person reaching Salesforce from Slack, Claude or Gemini Enterprise is still a user and still needs a license, and a license usable from every surface is easier to justify. The pressure on the price list comes from somewhere else: the job-ready agents launched the same week, which take on work that used to require a licensed person. Each task they absorb is a task no seat is needed for, and the Agentic Work Unit is not yet an answer. AWUs count “a prompt processed, a reasoning chain completed, or … a tool invoked” as the same unit, whether they are equivalent or not." ]]> tag:dlvr.it,2026-09-20:b5844a8812cef3d04e08039304e9b112 2026-09-20T16:04:25-07:00
"There is still an important distinction between action and accountability. The AI may be the actor. That doesn’t make human or institutional accountability disappear. We already understand this elsewhere: a police officer acts under laws, policies, training and delegated authority, and we still say the officer made the arrest. We don’t insist on saying that the institutional system caused the arrest. Agentic AI will increasingly work the same way." ]]> tag:dlvr.it,2026-09-20:37600e1a1329cf03363a80979a3702e9 2026-09-20T16:04:25-07:00
"Ask your SaaS vendor a straightforward question: what value am I getting from the products I’m already subscribed to and already paying for? Have them walk you through that value. Push them to be granular and specific. No canned marketing slides. This question accomplishes two things. First, it forces the SaaS vendor to account for the baseline. What is your current usage? What outcomes, ROI and/or benefits have the products delivered? Are you getting what was promised when you signed the contract? Do the fees paid align with the value received." ]]> tag:dlvr.it,2026-09-20:3524a3036b70b5110d19c4cb15456d68 2026-09-20T16:04:24-07:00
"All of this is an audacious viewpoint for a company like Salesforce, and it could speak to a future where the interface as we’ve known it melts away. It ties directly into the Headless 360 products announced earlier this year and the partnership with Anthropic around the recently announced Claudeforce product, each of which showed that Salesforce is walking the walk, not just talking the talk when it comes to eventually deprecating the interface. Headless gives customers a way to help agents connect to Salesforce data on the back end, while products like Claudeforce give them one way to replace the front end." ]]>