Daily Tech Digest by Kannan Subbiah: Daily Tech Digest

Quote for the day:

"Worry less, smile more. Don't regret, just learn and grow." -- @Pilotspeaker

The battle for agent connectivity: Can MCP survive the enterprise?

"MCP is the UI for agents. The future of asking to book an and have a pizza available when you arrive at the hotel only works if we have the connectivity," said Dag Calafell III, director of Technology Innovation at MCA Connect, an IT consultancy for manufacturers. But while seamless connectivity might be the Holy Grail for consumer apps, critics argue that it is irrelevant -- or even dangerous -- for the enterprise. ... Notably, MCP has significant backing from prominent companies, including Google, , Microsoft and its creator, Anthropic. Indeed, Calafell argued that while there are competitors out there, "MCP is winning" precisely because it has seen significant adoption by large software providers. Still, MCP clearly has significant issues -- mostly because it's in its infancy. MCP's rapidly evolving specification, uneven tooling, unclear security and governance controls, and lack of standardized memory, debugging, and orchestration make it better for experimentation than reliable enterprise use today. ... "There is room to innovate with a security-first 'MCP-like' standard that is resource aware, with trusted catalogues, privileges, scopes, etc. These would either be built on top of MCP, a sort of MCP v2, or introduced as part of a new protocol," said , co-founder and CTO at . And, of course, there remains an evolving trend that the AI industry will take an entirely different direction.

Digital Twin in Railways: A Practical Solution to Managing Complex Rail Systems

In the context of railways, are being deployed to improve asset lifecycle management, predictive maintenance, and infrastructure planning. By integrating inputs from IoT devices and advanced analytics platforms, these models help engineers monitor structural health, detect anomalies, and plan maintenance before failures occur. ... As the scale and complexity of rail networks continue to grow, the use of digital twins offers a unified, comprehensive view of interconnected assets, which empowers rail operators with faster decision-making and better coordination across departments. This technology is gradually becoming a core component of smart railway ecosystems. ... The architecture of a digital twin in railway systems is built upon the integration of multiple digital technologies, including Building Information Modelling (BIM), the Internet of Things (IoT), Geographic Information Systems (GIS), and data analytics platforms. Together, these technologies create a unified framework that connects the physical and digital environments of railway infrastructure and operations. ... The integration of operational data, including train movements, energy consumption, and passenger flows, allows operators to simulate different scenarios and optimise timetables, headways, and energy use. In dense networks such as urban metro systems, this contributes to improved punctuality and efficient energy utilisation.

Stop mimicking and start anchoring

It’s a fundamental truth that most are ignoring in their rush to emulate Big Tech playbooks. The result is a systematic misallocation of resources based on a fundamental misunderstanding of how value creation works across industries. ... the strategic value of IT should be measured by how effectively it addresses industry-specific value creation. Different industries have vastly different technology intensity and value-creation dynamics. In our view, CIOs must therefore resist trend-driven decisions and view IT investment through their industry’s value-creation to sharpen competitive edge. To understand why IT strategies diverge across industries shaped by sectoral realities and maturity differences, we need to examine how business models shape the role of technology. ... funding business outcomes rather than chasing technology fads is easier said than done. It’s difficult to unravel the maze created by the relentless march of technological hype versus the grounded reality of business. But the role of IT is not universal; its business relevance changes from one industry to another. ... Long-term value from emerging technologies comes from grounded application, not blind adoption. In the race to transform, the wisest CIOs will be those who understand that the best technology decisions are often the ones that honour, rather than abandon the fundamental nature of their business. The future belongs not to those who adopt the most tech, but to those who adopt the right tech for the right reasons.

Build vs buy is dead — AI just killed it

Ssomething fundamental has changed: AI has made building accessible to everyone. What used to take weeks now takes hours, and what used to require fluency in a programming language now requires fluency in plain English.When the cost and complexity of building collapse this dramatically, the old framework goes down with them. It’s not build versus buy anymore. It’s something stranger that we haven't quite found the right words for. ... And it's not some future state. This is already happening. Right now, somewhere, a customer rep is using AI to fix a product issue they spotted minutes ago. Somewhere else, a finance team is prototyping their own analytical tools because they've realized they can iterate faster than they can write up requirements for engineering. Somewhere, a team is realizing that the boundary between technical and non-technical was always more cultural than fundamental. The companies that embrace this shift will move faster and spend smarter. They’ll know their operations more deeply than any vendor ever could. They'll make fewer expensive mistakes, and buy better tools because they actually understand what makes tools good. The companies that stick to the old playbook will keep sitting through vendor pitches, nodding along at budget-friendly proposals. They’ll debate timelines, and keep mistaking professional decks for actual solutions. Until someone on their own team pops open their laptop, says, “I built a version of this last night. Want to check it out?,”

Quantum Tech Hits Its “Transistor Moment,” Scientists Say

“This transformative moment in quantum technology is reminiscent of the transistor’s earliest days,” said lead author David Awschalom, the Liew Family Professor of molecular engineering and physics at the University of Chicago, and director of the Chicago Quantum Exchange and the Chicago Quantum Institute. “The foundational physics concepts are established, functional systems exist, and now we must nurture the partnerships and coordinated efforts necessary to achieve the technology’s full, utility-scale potential. How will we meet the challenges of scaling and modular quantum architectures?” ... Although advanced prototypes have demonstrated system operation and public cloud access, their raw performance remains early in development. For example, many meaningful applications, including large-scale quantum chemistry simulations, could require millions of physical qubits with error performance far beyond what is technologically viable today. ... “While semiconductor chips in the 1970s were TLR-9 for that time, they could do very little compared with today’s advanced integrated circuits,” he said. “Similarly, a high TRL for quantum technologies today does not indicate that the end goal has been achieved, nor does it indicate that the science is done and only engineering remains. Rather, it reflects a significant, yet relatively modest, system-level demonstration has been achieved—one that still must be substantially improved and scaled to realize the full promise.”

Before you build your first enterprise AI app

Model weights are becoming undifferentiated heavy lifting, the boring infrastructure that everyone needs but no one wants to manage. Whether you use Anthropic, OpenAI, or an open weights model like , you are getting a level of intelligence that is good enough for 90% of enterprise tasks. The differences are marginal for a first version. The “best” model is usually just the one you can actually access securely and reliably. ... We used to obsess over the massive cost of training models. But for the enterprise, that is largely irrelevant. AI is all about inference now, or the application of knowledge to power applications. In other words, AI will become truly useful within the enterprise as we apply models to governed enterprise data. The best place to build up your AI muscle isn’t with some moonshot agentic system. It’s a simple retrieval-augmented generation (RAG) pipeline. What does this mean in practice? Find a corpus of boring, messy documents, such as HR policies, technical documentation, or customer support logs, and build a system that allows a user to ask a question and get an answer based only on that data. This forces you to solve the hard problems that actually build a moat for your company. ... When you build your first application, design it to keep the human in the loop. Don’t try to automate the entire process. Use the AI to generate the first draft of a report or the first pass at a SQL query, and then force a human to review and execute it.

Cloudflare reveals AI surge & Internet ‘bot wars’ in 2025

Cloudflare reported that use of AI models and AI crawling activity increased sharply. It said crawling for model training accounted for the majority of AI crawler traffic during the year. Training-related crawlers generated traffic that reached as much as seven to eight times the level of retrieval-augmented generation and search crawlers at peak. Traffic from training crawlers was also as much as 25 times higher than AI crawlers tied to direct user actions. The company said was the most widely used on its network. It was used by more than three times as many accounts as the next most popular models from providers such as OpenAI and . Cloudflare added that Google’s crawling bot remained the dominant automated actor on the Internet. It said ’s crawl volume exceeded that of all other leading AI bots by a wide margin and was the largest single source of automated traffic it observed. ... Cloudflare reported a notable shift in the sectors that face the highest volume of . Civil society and non-profit organisations became the most attacked group for the first time. The company linked this trend to the sensitivity and financial value of the data held by such organisations. This includes personal information about donors, volunteers and beneficiaries. Cloudflare’s data also showed changes in the causes of major Internet outages.

Who Owns AI Risk? Why Governance Begins with Architecture

But as AI systems grow more complex, so do their risks. Bias, opacity, data misuse, model drift, or even overreliance on AI outputs can all cause serious business, ethical, and reputational damage. This raises an uncomfortable question: who actually owns the risk of AI? ... AI doesn’t live in isolation. It consumes enterprise data, depends on cloud services, interacts with APIs, and influences real business processes.Governance, therefore, can’t rely on policies alone, it must be designed, structured, and embedded into the architecture itself. For instance, companies like Microsoft and Google have embedded AI governance directly into their architectural blueprints creating internal AI Ethics and Risk Committees that review model design before deployment. This proactive structure ensures compliance and builds trust long before a model reaches production. ... In other words, AI Governance is not a department, it’s an ecosystem of shared responsibility.Enterprise Architects connect the dots, Business Owners set the direction, Data Scientists implement, and Governance Boards oversee. But the real maturity comes when everyone in the organization, from the C-suite to the operational level, understands that AI is a shared asset and a shared risk. ... Modern enterprise architecture is no longer only about connecting systems. It’s about connecting responsibility. The moment artificial intelligence becomes part of the business fabric, architecture must evolve to ensure that governance isn’t something external or reactive, it’s embedded in the very design of every AI-enabled solution.

The 5 power skills every CISO needs to master in the AI era

According to the , nearly 40% of core job skills will change by 2030, driven primarily by AI, data and automation. For security professionals, this means that expertise in network defense, forensics and patching — while still essential — is no longer enough to create value. The real impact comes from how we interpret, communicate and apply what AI enables. ... The biggest myth in security is that technical mastery equals longevity. In truth, the more we automate, the more we value human differentiation. Success in the next decade won’t depend on how much code you can write — but on how effectively you can connect, translate and lead across systems and silos. When I look at the most resilient organizations today, they share one trait: They see cybersecurity not as a control function, but as a strategic enabler. And their leaders? They’re fluent in both algorithms and empathy. The future of cybersecurity belongs to those who build bridges — not just firewalls. Cybersecurity is no longer a war between humans and machines — it’s a collaboration between both. The organizations that succeed will be the ones that combine AI’s precision with human empathy and creative foresight. As AI handles scale, leaders must handle meaning. And that’s the true essence of power skills. The future of cybersecurity belongs to those who can blend AI’s precision with human expertise — and lead with both.

Manufacturing is becoming a test bed for ransomware shifts

“Manufacturing depends on interconnected systems where even brief downtime can stop production and ripple across supply chains,” said , Director of Threat Research, . “Attackers exploit this pressure: despite encryption rates falling to 40%, the median ransom paid still reached $1 million. While half of manufacturers stopped attacks before encryption, recovery costs average $1.3 million and leadership stress remains high. Layered defenses, continuous visibility, and well-rehearsed response plans are essential to reduce both operational impact and financial risk,” Rose continued. Teams were able to stop attacks before encryption in a larger share of cases, which likely contributed to the decline. Early detection helped reduce disruption, although strong detection did not guarantee a smooth recovery. ... IT and security leaders in manufacturing see progress in some areas but ongoing gaps in others. Detection appears to be improving. Recovery is becoming steadier. Payment rates are declining. But operational weaknesses persist. Skills shortages, aging protections, and limited visibility into vulnerabilities continue to contribute to compromises. These factors shape outcomes as much as attacker capability. The findings also show a need for stronger internal support. Security teams are absorbing organizational and emotional strain that can affect long term performance. Manufacturing operations depend on stable systems, and teams cannot maintain stability without workloads they can manage.

Daily Tech Digest by Kannan Subbiah

Daily Tech Digest - December 16, 2025