Category hub
Verify the person, not the voice or video
Synthetic media can imitate people you trust. Use a second channel and a pre-agreed verification phrase before acting.
Deepfake & AI Voice/Video Fraud
Synthetic audio and video used to impersonate a relative in distress, an executive authorising a transfer, a celebrity endorsing an investment, or to create non-consensual explicit imagery for extortion.
AI-Enabled Threats · Recognise, prevent, report and recover →
AI-Generated Phishing & Scam Content
Generative models producing fluent, personalised phishing, scam scripts and fake websites in any language at unlimited scale — removing every language and quality signal awareness training used to rely on.
AI-Enabled Threats · Recognise, prevent, report and recover →
Prompt Injection & LLM Abuse
Hostile instructions hidden in content that an AI system reads, causing it to leak data, take unauthorised actions or bypass its guardrails — the defining vulnerability class of LLM-based applications.
AI-Enabled Threats · Recognise, prevent, report and recover →
Synthetic Identity Fraud
Fabricated identities built from a mix of real and invented data — increasingly with AI-generated faces and documents — used to pass KYC, build credit history, and disappear with the balance.
AI-Enabled Threats · Recognise, prevent, report and recover →
Training Data & Model Poisoning
Corruption of the data or model artefacts used to train or fine-tune AI systems, creating backdoors or systematic bias.
AI-Enabled Threats · Recognise, prevent, report and recover →