Engineering Software Licensing

AI is disrupting a lot of industries. Music licensing is one of them. I was on the phone with a large enterprise customer who was planning to churn. They love our music and our service. But their team wanted to switch to an AI music generation tool. So I asked a simple question: “Have you reviewed the licensing implications for commercial use?” Silence. I followed up: Do you know how the model was trained? Do you have indemnification if there is a copyright claim? Are you protected if the output resembles someone else’s work? To their credit, they paused and brought in legal. A few days later, they called back and decided to stay. Not because AI is bad. But because they realized they had not fully evaluated the risk. AI is powerful. But in enterprise environments, power without protection becomes liability. If you are experimenting with AI-generated music or content, make sure you understand: -Training data exposure -Commercial usage rights -Indemnification coverage Innovation matters. So does due diligence. The companies that win will take both seriously. Our commercial-use music library is built for enterprise protection with clear rights, real artists, and real indemnification. Check it out: https://lnkd.in/efWhpP7U #AI #EnterpriseLeadership #MarketingStrategy #Copyright #RiskManagement #MusicIndustry #Innovation

Over the past few weeks, several CEOs of large, reputable companies have publicly claimed that large enterprises shouldn't work with frontier LLM providers because those providers will use your data to compete against you. Comments like these, which are neither particularly helpful nor accurate, create a lot of unnecessary confusion. So let me set the record straight. If an enterprise lets employees use consumer versions of frontier LLMs, then yes: those providers generally can, and by default do, use your data to improve their models unless you explicitly opt out. That's a real risk, and a good reason to govern which tools employees use for company work. But that is not the enterprise story. The three leading frontier providers (Google, OpenAI, and Anthropic) all offer enterprise agreements with standard terms that prohibit training on, or repurposing, data from their business customers. Here is representative language, taken verbatim from the publicly available commercial terms of one of the three: Customer Content. As between the parties and to the extent permitted by applicable law, [Provider] agrees that Customer (a) retains all rights to its Inputs, and (b) owns its Outputs. [Provider] disclaims any rights it receives to the Customer Content under this Agreement. [Provider] hereby assigns to Customer its right, title and interest (if any) in and to Outputs. [Provider] may not train models on Customer Content from Services. [Provider] will only access or use Customer Content to (i) provide the Services to Customer, (ii) comply with applicable laws, or (iii) enforce this Agreement. This is very consistent with the terms you would find from any hyperscaler. In many ways, frontier providers are a new kind of hyperscaler. For anyone looking to compete with the frontier labs: compete on price, features, and quality, not on sowing inaccurate stories about them.

🚨 New Zealand publishes its AI Strategy and Responsible AI Guidance, including how to 👉 ethically source datasets from a copyright perspective. World, take note: 1. Directly license copyright works: "AI developers are increasingly striking partnerships with traditional publishers and media entities to license their extensive content libraries, in order to foster innovation and grow together. Consider creating opportunities to partner directly with media libraries, publishers, iwi, and other content creators, rightsholders and aggregators." 2. Access a collective license "AI developers can access traditional ways to licence use of copyright works through collective licensing schemes offered by various copyright management organisations. Collective licences can also be used to obtain permission to use overseas works, vastly increasing the available volume and variety of copyright works available to AI developers. For New Zealand copyright works and business licence solutions, Copyright Licensing New Zealand intends to release a collective licensing scheme later in 2025 to partner AI developers with New Zealand rightsholders." 3. Use Fair Marketplaces "New marketplaces are emerging for creators and rightsholders to directly license their creative works for AI training, ensuring permission and remuneration. US examples include Created by Humans and RHEI." 4. Choose a fairly trained and commercially safe AI model "Examples of AI models that exemplify trustworthy AI and exclusively use licensed datasets include, but are not limited to: - Te Hiku  - Pro Rata AI - Adobe Firefly Video Model You can also check the internet for reporting on AI model infringement claims and infringement checking tools are also emerging." - 👉 Download the two documents below. 👉 NEVER MISS my essays and curations: sign up for my newsletter below (67,500+ subscribers).

𝗜𝗖𝗦 𝗔𝗰𝗰𝗲𝘀𝘀 𝗖𝗼𝗻𝘁𝗿𝗼𝗹: 𝗞𝗲𝗲𝗽𝗶𝗻𝗴 𝗖𝘆𝗯𝗲𝗿 𝗧𝗵𝗿𝗲𝗮𝘁𝘀 𝗢𝘂𝘁 𝟯:𝟬𝟬 𝗮.𝗺. 𝗶𝗻 𝗮𝗻 𝗲𝗻𝗲𝗿𝗴𝘆 𝗽𝗹𝗮𝗻𝘁: An operator sees the cursor moving—on its own. In 2021, hackers actually took control of a Florida water plant, nearly poisoning the water. Why? Shared passwords and open remote access. Access control in Industrial Control Systems (ICS) isn’t just IT hygiene—it’s a frontline defense. Unlike IT, ICS must balance security vs. uptime, making access control complex. 𝗞𝗲𝘆 𝗖𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲𝘀 𝗶𝗻 𝗜𝗖𝗦 𝗔𝗰𝗰𝗲𝘀𝘀 𝗖𝗼𝗻𝘁𝗿𝗼𝗹 ❌ Default & Shared Credentials – Many OT devices still use factory-set or hardcoded passwords. ❌ Overprivileged Accounts – Admins using the same account for both daily tasks & critical operations. ❌ Uncontrolled Remote Access – Unrestricted RDP, TeamViewer, or VPN access directly into OT. ❌ Lack of Continuous Audits – Old user accounts lingering long after employees leave. 𝗣𝗿𝗮𝗰𝘁𝗶𝗰𝗮𝗹 𝗦𝗼𝗹𝘂𝘁𝗶𝗼𝗻𝘀 (Aligned with IEC 62443) ✏️ Kill Default Credentials – Change all default passwords before deployment. Use compensating controls if you can’t. ✏️ Unique, Least-Privilege Accounts – No shared logins. Admins should have separate work and privileged accounts. ✏️ Secure Remote Access – Jump servers, MFA, and firewalls between IT & OT. No direct access to controllers. ✏️ Regular Audits & Offboarding – Disable accounts immediately when employees or contractors leave. 𝙍𝙚𝙘𝙚𝙣𝙩 𝙇𝙚𝙨𝙨𝙤𝙣: The Florida water plant breach could have been prevented with MFA, segmented access, and unique passwords. Simple steps can block attackers from turning small mistakes into disasters. ICS security is about access—who gets in, what they can do, and when they’re removed. Every login should tell a secure story. #ICS #CyberSecurity #IEC62443 #AccessControl #OTSecurity

AI Builder in Power Automate is an awesome tool. Between document processing, image and text analysis and text generation with prompts, it makes it very easy to build intelligent automation solutions. However, like all great things, it comes at a cost. But unlike most other awesome Power Platform tools, it can actually be quite expensive. The thing is that AI Builder billing is based on consumption. Each operation consumes a certain number of 'credits' depending on model and operation. For example, based on the latest version of the licensing guide, document processing using a custom model will consume 100 credits per document page. Invoice/PO processing using a pre-build model will consume 32 credits per document page. This already indicates it is usually best to use existing pre-built models when applicable, even if you may sometimes get the urge to train your own model, since it is so easy to do. Now, we get some of those credits for certain paid Power Platform licenses we purchase. For example, we get 5k credits per month for each paid Power Automate license (regardless of which type), or 500 credits per month for each Power Apps Premium license. So, with a Power Automate Premium license for $15/month, we could process 50 document pages using a custom model, or ~156 invoice pages using a pre-built model. This is usually not enough for a proper use case, except, maybe in some micro businesses. So, the alternative is purchasing a dedicated AI builder credit capacity add-on. But this bad boy comes in batches of 1M credits for $500/month. Could easily be overkill for most SMBs. And there is nothing in between, really. No pay-as-you-go plan either. On top of that, credits that are not consumed within a month do not roll over to the next month. But what quite a few people don't know is that the credits received via premium licenses can be pooled at tenant level for up to 1M credits per license type. So, we could technically purchase 200 Power Automate Premium licenses and get those same 1M credits. This obviously doesn't make sense, as 200 PA Premium licenses would set you back $3.000, which is 6x more expensive than the add-on. But, if you do not actually need the 1M credits, you could purchase premium licenses to a point that covers your needs. It technically makes more sense to purchase up to 33 Power Automate Premium licenses than a single AI Builder capacity add-on, if 165k credits is enough for you. And if anything less than that is enough, it is a no-brainer to go with a bunch of premium licenses instead of the add-on. And on top of that - you also get premium licenses for your team, which allow them to build amazing things with Power Automate. So, you get more value for less money. But if you need more than the 165k, then the add-on actually makes sense. And at that level, the cost should actually also be quite reasonable, as the benefits gained by automating over 5k documents per month should pay back for the $500.

Today's AI Mindset newsletter! This week revealed a fundamental conflict in AI development- Child safety law for chatbots! (Monday). Adult content on ChatGPT! (Tuesday). This shows AI companies can't serve both consumers (fewer restrictions) and enterprises (control and zero liability) with the same product. What it means for your business. Here's your TL;DR... (BUT FIRST: Subscribe at ai-mindset dot ai and get the full newsletters every week!) +++++ Why This Matters for Business: Trust can be an adoption killer. When AI brands chase both enterprise control and consumer freedom, nobody trusts the tools. Leadership sees liability nightmares, IT can't guarantee governance, and employees freeze rather than risk using the "wrong" AI The perception problem is real: Your ChatGPT Enterprise may be locked down, but when the same brand makes headlines for adult content, good luck explaining that distinction to your board, customers, or regulators The compliance gap widens: Just as companies are figuring out basic AI governance, the goalposts move again. California's safety law adds new requirements while OpenAI loosens restrictions, creating a regulatory minefield Hidden productivity costs: Confused employees quietly avoid AI altogether rather than risk violations, leaving you with expensive licenses and zero ROI What Companies Should Do Now: - Create crystal-clear dual policies: List approved enterprise tools (Copilot, Gemini Enterprise, Claude for Work) vs. consumer AI. Be explicit: company data never touches consumer tools, period - Six-week migration deadline: Before OpenAI's December changes, audit who has access to what, migrate all consumer accounts to enterprise versions, and implement data loss prevention rules - Turn transparency into trust: Don't just announce rules. You have to show teams the headlines, explain the business risks, and focus on what they CAN do with AI, not just restrictions - Vet vendors like investments: Ask where they're investing (enterprise governance or consumer engagement?), demand advance notice of policy changes, and prioritize partners who understand that boring enterprise features pay the bills December's changes make this week's confusion look quaint. Act now or explain later why your AI investment became a compliance nightmare. +++++++++ UPSKILL YOUR ORGANIZATION: When your organization is ready to create an AI-powered culture—not just add tools—AI Mindset can help. We drive behavioral transformation at scale through a powerful new digital course and enterprise partnership. DM me, or check out our website.

Selling to ENT without changing your pricing model is like showing up to a black-tie event in flip flops. MM pricing models don’t survive in enterprise sales. Why? Because selling 1,000 licenses to an enterprise isn’t 20x harder than selling 50 - but if you don’t adjust your pricing strategy, it will be 20x more painful. Enterprise buyers don’t think in per user terms. They think in budgets, forecasts, and cost centers. They want predictability, not a CPQ nightmare where they’re adjusting seat counts every quarter. If you’re moving upmarket, here’s how to avoid looking like a tourist at the grown-ups’ table: 1. Kill per-user pricing for large accounts. Enterprise CFOs see per-user models as a ticking time bomb...every new hire adds cost. Instead, sell in committed tiers, annual volume contracts, or all-you-can-eat licenses. - Instead of “$50 per user, per month,” structure it as, “$X for up to 1,000 users.” - Price for usage, not headcount - think storage, API calls, transactions, etc. 2. Enterprise doesn’t “expand naturally.” Build in expansion from day one. For MM, you can land small and grow. Enterprise doesn’t work that way. - Ramp pricing: Year 1 at 60%, Year 2 at 80%, Year 3 at 100%. Predictable growth, no CFO freak-outs. - Auto-expansion clauses: If usage exceeds X%, licenses auto-scale. Protects you from procurement pulling a “we’ll just add seats later” stunt. 3. Enterprise buyers expect to “win.” Give them a win - without losing. These buyers are trained to negotiate. They want a lower per-unit cost, but they’ll commit bigger dollars to get it. - Introduce an ENT Rate...lower per-unit cost, but higher minimum commit. CFOs love “efficiency,” and you get more ARR locked in. - Structure custom packaging that makes them feel special. Limited access to beta features, priority support, or bundled services. Want to win in enterprise? Stop selling like an SMB rep. Price for scale, control the expansion, and let procurement “win” on terms that make your CFO smile.

I didn’t learn SAM. I lived it. And so can you. When I started my career in software sales, I had no idea that one day I'd be helping enterprises save millions through SAM. My next step into software audits introduced me to the complex world of licensing — Microsoft, Adobe, IBM, Oracle — you name it, I had to learn it. From there, I got the opportunity to work with Snow, Flexera, and finally ServiceNow — where the real learning began. Back then, SAM Pro on ServiceNow was new. Documentation was limited, best practices were still forming, and most people didn’t know where to begin. I struggled too. But self-learning changed everything. Here’s the good news — today, the path is much clearer. If you're starting your ServiceNow SAM journey, here’s exactly how you can do it the smart way: Step-by-Step Guide to Learn ServiceNow SAM Pro (Hands-on) 1. Launch your Personal Developer Instance (PDI) Go to the ServiceNow Developer Site, request a PDI, and activate the following plugins: -Software Asset Management (SAM) Professional -Software Asset Management Foundation -Discovery with Demo Data This gives you a full-fledged playground with enough data to explore licensing, installations, users, and software models. 2. Navigate Through the SAM Application (Explore like a detective) Explore each section to understand what data is displayed and where it comes from: - Software Models - Discovery Models - Installations - Entitlements - Software Reconciliation Results Understand the tables behind each view: - cmdb_sam_sw_install → Software Installations - alm_license → Software Entitlements - samp_software_model → Software Models 3. Clean Up the Demo Entitlements - Delete all existing demo entitlements and start fresh. - Pick a vendor like Microsoft or Adobe — both offer well-documented licensing models that are easy to begin with. 4. Create New Software Entitlements Define new license types manually. Set the appropriate metric types (Named User, Device, Per Core, etc.). Map entitlements to discovery models and installations, then generate a compliance report. This is where you'll begin to understand: - How SAM calculates compliance - How installations match entitlements - How software usage drives licensing 5. Practice Cloud License Allocation (Adobe Use Case) Delete demo users tied to Adobe Creative Cloud in your PDI. Then: - Create new users - Allocate cloud licenses - Observe how ServiceNow calculates consumption This step helps you understand the difference between installed software and cloud-based subscriptions. Rest In the Comment Section. #ServiceNow #SAMPro #ITAssetManagement #LearningJourney #SelfLearning #SAM #ServiceNowCommunity #CareerGrowth #Motivation