Technology

Financial crime compliance may be enterprise AI's toughest proving ground. A model can be accurate and still be unusable if a compliance team cannot explain, govern, or audit its output. One interesting chat I had recently on this topic was with the Flagright team. They are building an AI operating system for financial crime compliance that brings  transaction monitoring, watchlist screening, risk scoring, case management, investigations, regulatory filing, and governance into one system. What I found interesting about their approach: ◾AI is embedded in real investigation workflows, not bolted on as a separate chatbot. ◾Teams can surface relevant evidence, investigate alerts, draft narratives, and recommend next steps while keeping human oversight. ◾Decisions remain explainable and audit-ready. ◾One operating layer reduces fragmented tools and handoffs. I think some of this thinking applies well beyond financial services too. AI in production does not just need to work. It also needs to be governed from the start and accountable when decisions matter. And congrats to the Flagright team on their $12.5M Series A, led by Infinity Ventures. 📍Worth exploring: https://lnkd.in/gGJAU7u6 #AI #FinancialCrime #Compliance

IKEA automated the task, not the people. I have always believed that the purpose of automation should be to remove repetitive work, but IKEA’s approach shows what that principle looks like when a company applies it at scale. Its AI chatbot, Billie, began handling routine customer enquiries, which could easily have turned into a familiar cost-cutting exercise. Instead, IKEA reskilled 8,500 call-centre employees in areas such as remote interior design, digital sales, relationship-building, and more complex customer support. What I find especially interesting is that IKEA did not treat those employees as redundant simply because part of their work had become automatable. It looked at the knowledge they had already built about customers and products, then asked how that experience could create more value elsewhere. The wider remote sales channel generated €1.3 billion in revenue in 2022. We cannot attribute that entire figure to reskilling alone, but the strategic direction still matters. In my Human+AI Equation, technology provides scale and speed, while people provide judgment, creativity, and connection. IKEA did not choose between the two. It redesigned the work around both. The best automation strategy does not begin by asking, “How many people can we remove?” It begins by asking, “What more valuable work can our people do now?” How would your AI strategy change if every automation plan also required a reskilling plan? #HumanAgentOrchestrator #AITransformation #HybridManagement #WorkforceReskilling

Solar created the duck curve. Batteries are now flattening it. This chart shows median wholesale electricity prices in New South Wales during the second quarter of each year. In 2023, prices surged towards $300/MWh during the evening peak. In 2026, that peak is below $100/MWh. While the chart shows NSW, this same pattern is emerging across Queensland, Victoria and South Australia as well. The driver of this is becoming familiar: batteries. ✅ Batteries create demand during solar-rich daytime hours, preventing midday prices from falling as deeply. ✅ They discharge after sunset, adding supply when evening demand rises. ✅ This reduces reliance on gas and hydro during the evening peak, when they have traditionally set higher prices. The scale of the battery build-out is now large enough to reshape both electricity supply and demand. Modo estimates that registered home-battery power capacity across Australia's National Electricity Market has reached 7.8 GW, overtaking the 7.7 GW of operational grid-scale batteries. Those home batteries increasingly charge during solar-rich daytime hours, then meet household demand after sunset - reducing grid demand. Meanwhile, grid-scale batteries add supply directly into the evening market. Batteries are also increasingly determining the wholesale price. In Q2, battery charging and discharging set prices in 36% of NEM dispatch intervals, up from 17% a year earlier. During the evening peak, battery discharge was the dominant price-setter. The curve is therefore being flattened from both directions: more demand when solar is abundant and more supply when evening demand rises. As battery capacity grows, the gap between the cheapest and most expensive hours is narrowing. That means lower peak prices, less volatility and a more productive use of daytime solar.

Our frontend builds crossed 20 minutes. The obvious move was to rip out Webpack for something faster. We didn't. Before touching a single line of config, we profiled the entire pipeline. Webpack's ProgressPlugin turned up something we didn't expect — one plugin, Terser, was eating 13+ minutes of every build. Webpack wasn't the bottleneck. JavaScript minification was. 🔍 That one number changed the whole plan. We kept Webpack. Replaced Babel with SWC. Swapped Terser for esbuild. Moved CSS optimisation to LightningCSS. Parallelised compilation across cores with thread-loader. ⚙️ Fixed a hidden bottleneck too — AWS CodeArtifact tokens were re-authenticating every 12 hours and quietly busting our Docker cache. Restructured the Dockerfile into layers that actually cache. 🐳 20 minutes to 2 minutes. Cold build ~90% faster. Incremental ~95% faster. Across 20 production apps, 600+ PRs a month, 5,000+ CI runs. ⚡ No migration. No four-week rewrite. No new bundler to debug for the next year. The real lesson wasn't about tooling. It was that we almost skipped measurement and went straight to migration — the industry default. Glad we didn't. 💡 Full breakdown — the Speed Measure Plugin data, the pnpm switch, the exact Dockerfile changes — link in the first comment. 👇 Really proud of what our engineering team pulled off here. Worth the read. 🫡 Zepto Zepto Tech #ZeptoTechXPress #ZeptoEngineering

Every January, the same headlines. EV sales are slumping. Nobody wants electric cars. The bubble has burst. And every year, the data tells a different story. In 2015, battery electric cars made up 0.3% of new UK registrations. The 12-month average now stands at 25.1%. One in four new cars registered in the UK is fully electric. So why do the collapse stories keep coming? Because of a predictable seasonal pattern. Manufacturers push EV registrations into December to meet fleet targets and the ZEV mandate. January then looks weak by comparison. It happens every single year. The spike and the dip are two halves of the same mechanism. But: the December peaks are becoming less exceptional. In 2022, December’s EV share stood 16 percentage points above that year’s average. In 2025, just 6 points. The baseline has caught up with the peaks. And June 2026 came in at 29.9%. Within touching distance of December levels, in an ordinary month. The rolling average has climbed for a decade without a single sustained decline. You can question the pace as being too slow. But you cannot question the direction of travel as the data is pretty clear. Next time you read that nobody wants EVs, check the trend line over the last decade rather than a single month. Data: DfT vehicle licensing statistics (2015 to 2025), New AutoMotive (2026).

Of course you need to use open-source models if you’re an enterprise leader. Close model providers, that are now forcing data retention, are gaining immense leverage on your business if you don’t. As you connect models to your business context, they see it and learn from it, and have a track record of going after their most successful customers thanks to this information. But that’s not enough, you also need to store your data and records in open systems, or your software vendors might block you from building AI systems outside of the walled garden they have set up for you. If you can’t convince them to give you complete access to the data they manage for you, AI fortunately allows you to migrate quite fast. Once you’ve got hold of your data, you’ll need to manage how AI systems can access this data on behalf of human users, because you don’t always want Bob to see what Alice is doing in your company. That’s hard and merciless, since AI models are great at finding need-to-know errors. It takes systems that check hard access rules and models that check soft access rules. Now comes the most important part. You need to set up your own continuous training flywheel, so that you can improve your AI systems based on their interaction with your employees and your users. This is how you turn the edges of your business into AI systems your vendors and competitors cannot replicate. It’s also how you reduce deployment cost as well, as you can shrink models according to model input distribution. Those bills are getting substantial, we need to collectively become efficient if we want AI development to continue, so that matters. All of these efforts might seem daunting – they are. This is both a complete replatforming of your IT, and a complete change in the way you’re developing software, and operating your business. AI lifecycle management requires understanding human behavior and gradient descent, that’s a stretch. At Mistral, we facilitate that work by providing all primitives that you need in a single control plane, Studio, and a training platform, Forge. With our applied AI engineers and scientists working hand-in-hand with our customers, we ensure that we transfer knowledge, and that we can disappear once the systems are up and running. We deploy on our customers' infrastructure, or through our zero-data-retention hosted services, so that your edges remain your edges, and the switch button can be fully in your hand. Frontier AI can accelerate the growth of your business, but if it’s not in your hands, it’s not going to be your growth.

🚨 Zero Trust for AI Agents Anthropic just released "Zero Trust for AI Agents." As we're thinking about agentic permissions, applying a Zero Trust discipline is critical to secure adoption. AI agents interpret goals, call tools, chain actions, delegate to other agents, and maintain context across sessions. The trust surface is different. The paper introduces "least agency" — a concept that OWASP has been promoting — and the distinction from least privilege is worth sitting with. 👉 "Least privilege" asks what an identity can access. 👉 "Least agency" asks what an agent can do, under what conditions, with which tools, and with what level of oversight. Autonomous agents introduce action risk alongside access risk — and the boundaries around behavior need to be architecturally enforced, not assumed. The paper includes a design test worth writing down: 🔥 Does the control make the attack impossible, or merely tedious?🔥 The practical controls follow directly from Zero Trust fundamentals — cryptographic agent identity, short-lived credentials, tool allow-listing, sandboxed execution, and full traceability from prompt to action to outcome. None of this is new doctrine. It's existing architecture applied to a harder problem. Full disclosure: the paper cites NIST SP 800-207 on Zero Trust Architecture and the CISA Zero Trust Maturity Model, both of which I co-authored during my time supporting Federal Zero Trust efforts at CISA. Zero Trust is built for a world where we have to remove implicit trust. Agentic AI is the next version of that same problem — valid identities, valid credentials, legitimate-looking actions, and still no basis for assumed trust. Access is earned. Actions are constrained. Agency must be governed. 👉🏼 Link to Anthropic's paper in the comments.

Saudi Arabia built the world's largest virtual hospital, and we haven't even heard of it. It connects 224 hospitals and treats 400,000 patients a year without a single physical bed. It's called Seha Virtual Hospital in Riyadh, and it just earned a Guinness World Record for being the largest virtual healthcare provider in the world. But how can a hospital be “virtual”? How does it work? → Imagine you live in a small town with only a basic local hospital. → It has doctors and equipment. But if you need a cardiologist or neurologist, you travel 6+ to a bigger city. In urgent situations, people lose lives. → With Seha, specialists treat you remotely through your local hospital - reviewing scans, diagnosing conditions, prescribing treatment - while local staff execute it. That's the model. Specialist expertise delivered through existing hospitals. And here's what makes it work: ▶️ AI prioritizes urgent cases - analyzes CT scans and imaging to rank who needs immediate intervention ▶️ IoT monitors patients remotely - heart failure patients wear devices that alert doctors before hospitalization is needed ▶️ Integrated health records - manages prescriptions and reports across all 224 hospitals in real-time The results? - ICU patients now stay an average of 4 days instead of weeks.  - Stroke patients get CT scans within 25 minutes of arrival. - Treatment starts in 28 minutes.  - Radiology reports in 2 hours. This isn't telemedicine where you video-call a doctor from home. This is expertise delivered through your local hospital without the specialist being physically there. It proves you don't need cardiologists and neurologists in every town. You just need good internet and hospitals willing to collaborate. Do you think virtual hospitals could solve specialist shortages in rural areas? #Entrepreneurship #healthtech #innovation

Every AI company serving your org is currently losing money on you. The compute costs 10x what they're charging, which means we're all building on borrowed pricing. Claude pro costs $20/month. The compute to serve a power user runs $200-$400/month. Microsoft was losing $80/month on some Copilot users paying $10. What does that mean? They're buying market share with subsidized pricing. Your teams build AI into everything. It becomes infrastructure. Then prices go up and you can't leave. And this is already happening. GitHub moved Copilot to usage-based billing last month. OpenAI launched a $100/month tier. Anthropic's top plan is $200/month. Microsoft raised 365 prices twice in four years and pointed directly at AI costs. Meanwhile, most companies are still budgeting like $20/seat is the real price.... A 50-person team on Claude Pro costs $1,000/month on paper. At actual token rates, that same usage costs $15,000 to $40,000/month. Agentic AI made this worse. Coding agents don't use tokens like a chat conversation. They burn 10x more. GitHub literally said this is why they killed flat-rate pricing (!!). OpenAI and Anthropic are both heading toward IPOs. Public investors don't tolerate selling a dollar of compute for twelve cents. The moment these companies go public, the subsidies will likely die. Your finance team needs to audit actual token consumption across every team. Not seats. Not licenses. Actual usage. Then model what happens when prices go up 5x. The gap between what you pay today and what you'll pay in 18 months is going to be one of the ugliest budget surprises most companies have ever absorbed. Read the full piece on State of Brand here: https://lnkd.in/giCbt-HW

The AI Coding Revolution Is Here, But Are We Testing for It? As AI-assisted development reshapes how we build software, I've been thinking a lot about something that is talked about often but doesn't always get the focus it deserves: automated testing. At JPMorganChase, we're embracing AI coding tools to accelerate delivery, reduce toil, and empower our teams to focus on the work that matters, reducing cognitive load of repetitive tasks. But speed without safety is just risk in disguise. Here's what I believe every leader (and this is broader than technology) needs to consider right now: • AI writes code faster than humans can review it manually. If your testing strategy is still largely manual, you're already behind. AI-generated code can introduce subtle logic errors, security vulnerabilities, or edge-case failures that look perfectly reasonable on the surface. Automated testing is no longer a best practice, it's a non-negotiable safeguard. • Test coverage is your new quality contract. When AI is your co-developer, the test suite becomes the specification. If you can't describe expected behavior in a test, you can't trust what the AI builds. Investing in robust unit, integration, and regression testing frameworks is investing in the integrity of your entire delivery pipeline. • Shift-left testing amplifies AI's value. It doesn't slow it down. Some worry that rigorous testing will negate the speed gains from AI coding. The opposite is true. When automated tests are embedded early in the development lifecycle, AI tools can iterate faster, self-correct, and validate outputs in real time. Testing enables velocity; it doesn't constrain it. • Your teams need to evolve alongside the tools. The best teams of tomorrow won't just write code. They'll architect test strategies, evaluate AI outputs critically, and build systems that are observable and verifiable by design. We owe it to our teams to invest in this skill evolution now. At the scale we operate, serving millions of customers, the cost of a defect isn't just technical. It's trust. And trust, once broken, is hard to rebuild. AI is a force multiplier. But multiplying without a strong foundation multiplies risk just as fast as it multiplies output. Build fast. Test smarter. Ship with confidence. I'd love to hear how other leaders are thinking about quality engineering in the age of AI. What's working for your teams? #AIEngineering #SoftwareTesting