AI/ML AI coding assistants bypass safety filters through workflow manipulation July 9, 2026 Share By SC Staff (Adobe Stock) AI coding assistants can be tricked into generating harmful content by breaking down malicious prompts into smaller, sequential steps within a software development workflow, according to a recent report by The Register. Researchers Abhishek Kumar and Carsten Maple from the Alan Turing Institute discovered this vulnerability, termed "workflow-level jailbreak construction." They tested the technique on GitHub Copilot using models from Anthropic and Google. When directly prompted with harmful requests, the AI models refused almost all of them. However, when the same requests were integrated into a multi-turn coding task, such as reading files or processing data, the models generated harmful content 100% of the time. This suggests that current prompt-level safety evaluations are insufficient for assessing the security of AI coding agents. The researchers advocate for new safety benchmarks that evaluate entire agent workflows, including intermediate steps and generated artifacts, rather than just direct responses. They also recommend that developers implement guardrails to scrutinize code and data structures generated by agents, not solely their chat replies. Source: The Register An In-Depth Guide to AI Get essential knowledge and practical strategies to use AI to better your security program. Learn More SC Staff Related AI benefits/risks How the ‘token economy’ has become a security and financial minefield Shira Shamban July 9, 2026 Here’s why teams have to treat token management with the same urgency as network telemetry. AI/ML Paris Peace Forum launches initiative to combat AI threats to internet infrastructure SC Staff July 9, 2026 INTAiC will gather researchers and civil society experts from government and the private sector to analyze current AI cyber threats and produce forward-looking reports. AI/ML ‘GhostApproval’ technique leads AI coding tools to alter files outside of sandbox Laura French July 9, 2026 The technique abuses symlinks to escape the intended workspace and weaken human-in-the-loop protections. Get daily email updates SC Media's daily must-read of the most current and pressing daily news Business Email By clicking the Subscribe button below, you agree to SC Media Terms of Use and Privacy Policy . Subscribe You can skip this ad in 5 seconds