News

More in News

• Google Chrome Fixes Over 1,000 Bugs Using AI Across Security Workflow

  Google has significantly integrated artificial intelligence (AI) into Chrome's security workflow, enabling faster discovery, triage, and patching of vulnerabilities. This strategic shift has allowed Chrome 149 and 150 to fix 1,072 security bugs, surpassing the total from the previous 23 stable releases combined.

  

• US Allies Update SBOM Guidance for Enhanced Software Security

  The US and 13 allied countries have issued updated guidance on Software Bill of Materials (SBOM) minimum elements, addressing new use cases and applications. This enhances software supply chain visibility, crucial for engineering teams to manage vulnerabilities and risks effectively.

  

• Oracle, Google Cloud Expand Partnership for Gemini AI Models in Enterprise Apps

  Oracle and Google Cloud have broadened their collaboration to integrate Google's Gemini AI models into Oracle's enterprise application portfolio. This move aims to provide developers and enterprises with more options for building AI agents and embedding AI capabilities directly into core business workflows.

  

• PyTorch Tutorial: Build Your First Digit Classifier from Scratch

  PyTorch, an open-source deep learning framework, has made building, training, and deploying AI models accessible to developers. This guide provides a step-by-step tutorial on constructing a handwritten digit classifier using PyTorch and PyCharm.

  

• DORA Research: AI Amplifies Software Development, Organisational Systems Key

  DORA's 2025 research on AI in software development highlights that AI acts as an amplifier, with the greatest returns stemming from a strategic focus on underlying organisational systems. Without this foundation, AI's productivity gains are often lost to downstream chaos, impacting overall delivery.

  

• Google AI fixes 1,072 Chrome security bugs, accelerates patching

  Google leveraged artificial intelligence (AI) to fix 1,072 security bugs in Chrome 149 and 150, exceeding the total fixes from the previous 23 releases combined. The company is integrating large language models (LLMs) across its vulnerability management process to enhance discovery and accelerate patch delivery.

  

• India IT Talent Demand Stabilises, Skill Gaps Persist for Freshers

  India's IT sector is showing signs of near-term talent demand stabilisation, with an estimated 109,000 job openings for August 2026, according to Xpheno. However, a significant gap in applied technical skills and problem-solving abilities among fresh graduates continues to challenge corporate readiness.

  

• Indian IT Fresher Hiring Shifts: AI Skills Now Crucial

  India's IT sector is undergoing a significant shift in fresher hiring, moving away from volume-based recruitment to a selective, demand-driven model. Major firms are prioritizing specialized skills in AI, data engineering, and cybersecurity, leading to delayed onboarding for many engineering graduates.

  

• Microsoft launches MAI-Cyber-1-Flash, Project Perception for AI patching

  Microsoft has introduced MAI-Cyber-1-Flash, its first in-house cybersecurity AI model, alongside Project Perception, an agentic system designed for automated vulnerability management. This initiative aims to find, triage, and patch security flaws efficiently.

  

• AI Agent Tool Access Demands Deterministic Policy, Not Just Prompts

  Securing AI agents that use tools requires a shift from relying solely on prompt engineering to implementing robust architectural controls. This ensures that tool access, which effectively becomes production access, is governed by deterministic policies and strict validation.

  

• Critical Ruflo Flaw Exposes AI Swarms to Unauthenticated RCE Attacks

  A critical vulnerability (CVE-2026-59726) in the open-source AI agent orchestration platform Ruflo allows unauthenticated remote code execution (RCE) in its MCP bridge container. This flaw enables attackers to gain shell access and tamper with AI outputs.