New research published in Science of Computer Programming tackles a critical challenge: securing the code that LLMs generate 🔒
The study by Aldawsari (2026) explores how prompt engineering techniques can be used to improve the security of LLM-generated code.
Kevin McPartland at Coalition Greenwich just published a piece that puts AI-assisted coding (or "vibe coding") squarely in the capital markets spotlight — and the examples he shares are worth paying attention to.
The shift from AI code completion to autonomous coding agents is accelerating fast 🚀
A recent piece on C# Corner maps out where AI-assisted programming is heading, and the trajectory is striking.
SD Times has published a piece looking at how organisations can avoid common pitfalls with AI coding assistants in 2026 — and how to actually extract value from them rather than just adding noise to the development process.
🔗...
OpenAI has launched a new company backed by $4 billion, dedicated to helping organisations build and deploy AI systems at enterprise scale.
That's a significant move.
IBM Consulting has launched Enterprise Advantage on AWS — what they're calling the industry's first enterprise-scale agentic AI platform natively integrated with AWS infrastructure. It's now generally available.
Citigroup just raised its AI market forecast to over $4 trillion, driven by enterprise adoption picking up pace 📈
The key driver? Commercial use cases like software development and agentic workflows that automate tasks across organisations.
MIT Sloan Executive Education just published a detailed look at how Merck KGaA, Darmstadt, Germany is approaching enterprise-wide AI adoption — and it's a genuinely useful case study for any executive thinking about scaling AI across their organisation 🔍
Here's what...
A striking stat from PwC data cited in this piece: just 20% of companies globally are capturing 74% of the total value generated by AI 📊
That's not a small gap.
Wharton management professor Rahul Kapoor makes a sharp argument in this piece: generative AI won't create business value on its own. The technology is an invention.
Software is becoming a commodity. And the numbers in this piece from M Accelerator make the shift feel very real 📊
Here's the headline: GitHub reports that 46% of all code is now AI-generated. Development velocity has jumped 55%+ for teams using AI tools.
AI coding assistants are reshaping how software gets built — and now they're disrupting how it gets priced, too 💡
Fast Company explores a growing challenge across the industry: as AI tools enable experienced developers (and even non-programmers) to produce working...
Six weeks. That's the average time we've watched enterprise teams shave off their delivery cycles this year — not by writing code faster, but by eliminating work that shouldn't exist in the first place 📉
Here's what we mean.
When most engineering organisations talk...
Forbes makes a sharp point here: the race to adopt AI across the enterprise is well underway, but speed alone won't determine who wins 🏁
The article highlights a real tension many organisations are facing right now.
Gartner projects that 80% of the software engineering workforce will need upskilling because of generative AI 📊
That's a staggering number.
The concept of an AI-native software development lifecycle is gaining real traction. EPAM NEORIS has been advancing what they call AI/Run SDLC initiatives — rethinking the entire development lifecycle with generative AI built in from the ground up, not bolted on as an...
Deloitte's 2026 State of AI in the Enterprise report is out, and it paints a detailed picture of where organisations actually stand with AI adoption right now 📊
Here's what stands out:
Two-thirds (66%) of organisations report productivity and efficiency gains from...
New research worth paying attention to if you're building with LLM-generated code 👀
A team of researchers has released LLMloop, a framework designed to automatically refine code and test cases produced by large language models through five iterative feedback loops.
Thrilled to welcome Highfield Qualifications to Upcast 🎉
Applied AI that delivers real value — not AI for its own sake. Excited to build this partnership.
highfieldqualifications.com | upcast.social
McKinsey found orgs with centralised AI governance are 2.5x more likely to see real productivity gains. The bottleneck isn't AI's power — it's invisible workflows and ungoverned adoption. 🤔 ability.ai/blog/ai-workflow-…