- Published
- Source
- CSO Online UK
- Filed under
- UK industry, Attacks, AI for defence
Although the Jev model rolled out last month by TypeSafe claims to have significant advantages over traditional LLMs, a security report published Wednesday argued that its streamlined approach makes it more vulnerable to some kinds of attack, compared to traditional LLMs. “In some cases, [Jev] is even more vulnerable to attacks, with both adversarial text attacks and prompt injection being able to disrupt the model’s outputs, resulting in incorrect and potentially devastating actions,” said the blog post from researchers at…
Using AI safely
- Know which AI tools your organisation uses, and what data goes into them.
- Treat AI agents like any other identity: least privilege, logging and a named owner.
- Test AI features for prompt injection and data leakage before customers reach them.
More helpNCSC: guidelines for secure AI system development (opens in a new tab)
Defender’s notebook
The attacker technique behind this story, in MITRE’s terms – and what to check in your own estate.
LLM Prompt Injection
ExecutionWhat a defender checks
- Treat everything an AI tool reads – email, web pages, documents – as untrusted input.
- Limit the data and tools an AI agent can reach, and give it its own least-privilege identity.
- Log prompts and tool calls into the SIEM so you can investigate.
Build the skillBlue Team Analyst · GenAI & AI Security Monitoring
© 2026 The MITRE Corporation. This work is reproduced and distributed with the permission of The MITRE Corporation. MITRE ATT&CK and ATT&CK are registered trademarks of The MITRE Corporation.
MyCyberWay training
AI security · in both programmes
AI is part of the attack surface now. Learn to watch it.
Blue Team Analyst covers Security Copilot in triage and monitoring the AI itself: inherited permissions, prompt injection and Purview DSPM for AI. Blue Team Engineer secures the deployment: Copilot pre-flight hardening, agent identity and tool allow-lists.
- Security Copilot in triage
- Prompt injection and DSPM for AI
- Agent identity and allow-lists
Headlines and short summaries only, except where a public body publishes under an open licence. Every story links to its publisher, who owns it. MyCyberWay does not edit or endorse third-party reporting.
