AppSec stories
Malicious packages can now be blocked before they reach developers' machines, as supply chain attacks shift into AI-assisted coding workflows.
Threats are now spanning identities, cloud and software supply chains, with AI speeding attacks and complicating defence, Microsoft says.
Mac developers using autonomous agents now get a free beta tool that blocks risky actions before tools, files or credentials are exposed.
The tool aims to close security gaps left by AI-written code before vulnerabilities reach security backlogs or pull requests.
Growing concern over autonomous AI systems has helped the Austin-based startup secure fresh backing to expand sales and revenue growth.
It aims to help engineers close software flaws before pull requests are opened, as AI coding tools accelerate release cycles and security backlogs grow.
Dozens of organisations worldwide face renewed compromise risk after attackers used an encoded path to slip past defences and hit Oracle PeopleSoft.
With most firms using AI but few staff ready, the course targets non-developers handling sensitive data and automations at work.
Security teams may cut backlogs as Wingman finds, fixes and verifies flaws inside AI coding tools before pull requests are opened.
Widening AI use across office teams is leaving many firms exposed to security and governance risks, the company says.
Attackers are exploiting AI faster than many defences can respond, prompting a shift to continuous testing for hidden exposures.
F5's recognition comes as enterprises scramble to secure AI agents and APIs, with 88 per cent reporting AI security challenges, its research found.
Software supply chain weaknesses are becoming the main entry point for attackers as development pipelines expand without enough oversight.
Autonomous systems are now creating fresh avenues for code theft, remote execution and runaway cloud spending, Mandiant says.
The stealer's use of typosquatted npm packages has raised fears that bug bounty channels are being abused to monetise stolen developer data.
With most firms using AI in daily work but few staff ready for it, the programme targets non-developers handling sensitive data and automations.
The private previews aim to help security teams spot exploitable code chains and connect AI agents to Rubrik's governance tools more safely.
Companies selling software into the bloc now have 24 hours to flag exploited flaws, with fines reaching EUR €15 million for delays.
Enterprise buyers are spending more to secure AI agents, with Snyk saying Evo now drives 60% of new deals and larger contracts.
Smartphone apps used for banking and workplace access will face tougher defences as the partners recruit two cybersecurity researchers for a 30-month project.
