Stack Overflow

AI agents are confused deputies, and they pose a real security threat when we give them the keys to our kingdoms. That's what happened with the recent Instagram breach, when hackers got control of more than 20,000 accounts using the Meta AI support agent. In this piece from The Heap, Fabio Salvadori explores what happened during that breach and why human judgement in our code is more important than ever. https://lnkd.in/eF2gm6Sv

🎙️ We welcome Trisha Gee, a Java champion and developer productivity advocate, to explore how AI is transforming the role of IDEs and the broader developer experience; the relevance of traditional tools, muscle memory, the risks of hype; and how to adapt workflows for AI-driven development. https://lnkd.in/eW8xUT2P

When AI gets things wrong, it’s usually down to three distinct flaws: hallucinations, sycophancy, and intentional deception. But luckily for AI developers, journalism solved these issues centuries ago. Ritoban Mukherjee explores how devs can translate 500 years of media frameworks into clear engineering solutions that can help prevent specific operational failures and make AI tools more accurate. https://lnkd.in/e2_NHqFi

🎙️ We welcome Bryan Clark, Director of Product for Lakebase at Databricks, to discuss what happens when AI agents become the primary creators and users of databases, why agents are “sloppy” about cleaning up infrastructure, and how database branching and centralized access control can help teams keep up with agent-driven development. https://lnkd.in/eC84yC_C

For daily users, single agent workflows seem to be the most comfortable for devs—giving them the most control while still increasing their productivity. 60% of respondents from our latest #StackOverflowKnows survey block agents from making unapproved system changes, and 69% prefer predictable, single-agent setups over complex, multi-agent configurations. Dive into the findings from our survey on agentic workflows: https://lnkd.in/eycTuVJ9

🎙️ On this episode of Leaders of Code, Eric Anderson, director of engineering at Intuit, joins our engineering director 👨💻 Ben Matthews to talk about what happens to software teams when AI makes code generation seemingly free, what it means for culture when product/engineering roles converge, and the skills that matter most in an AI-first industry. https://lnkd.in/eV7ehhTE

What if our AI agents could learn from each other? Introducing Stack Overflow for Agents, an API-first knowledge exchange built for the agentic era. No longer do you need to spend your time and tokens getting solutions that other agents have already figured out. Stack Overflow for Agents turns one agent’s expensive discovery into every agent’s reusable knowledge. Learn about Stack Overflow for Agents and join the beta now: https://lnkd.in/eQ4Q6DAH

🎙️Tanya Janca returns to the show, now part of the OWASP® Foundation Top 10 team, to discuss what changed in the latest OWASP Top 10 release, how the list shifted to a broader software supply chain focus, and why they added vibe-coding as awareness items. https://lnkd.in/e87uqsis

LLMs have become ubiquitous in the tech industry. But developers have favorites. According to our 2025 #DeveloperSurvey, Claude Sonnet (67.5%), Gemini Reasoning (65.2%), and OpenAI Reasoning (63.6%) are the models that devs are most satisfied with. After a year full of ever-evolving models, what do you think will be the highest rated LLM of 2026? Dive into more findings from our 2025 survey: https://lnkd.in/egtxEE76

🎙️ From the floor of HumanX, we welcome Songyee Yoon, managing partner at Principal Venture Partners (PVP), to chat about AI development outside the US, the need to adapt models to local languages and culture, and how venture capital is looking at international AI companies. https://lnkd.in/es-WNgXR