🔒 Staying Safe

AI-Powered Scams

AI has made scams more convincing. AI can clone voices, create realistic videos of people who don't exist, and write flawless phishing emails in any language. The safeguards that used to work — 'it sounded genuine' — are less reliable.

• Voice cloning: AI can replicate a voice from 3 seconds of audio.
• Deepfake video: realistic videos of public figures (or family members) saying things they never said.
• If something feels wrong, it might be — hang up, don't click, verify via a different channel.

---

Recognising AI Scam Calls

Fraudsters use AI to clone voices of family members ('grandparent scam') and to make automated calls sound personal. If a call asks for money urgently, hang up and call the real person back on a number you already have.

• The 'grandchild in trouble' scam: AI voice mimics your grandchild's distress to request urgent money.
• Always establish a family code word for genuine emergencies.
• Banks will never ask you to move money to a 'safe account' — this is always fraud.

---

AI and Email Phishing

AI has eliminated the spelling mistakes and awkward phrasing that used to signal phishing emails. Modern AI phishing is personalised, plausible, and polished. Judge emails by what they ask, not how well they're written.

• Red flags regardless of quality: urgency, unusual requests, requests for personal data.
• Hover over any link before clicking — the real destination often doesn't match the display text.
• If in doubt, don't click. Go directly to the organisation's website and log in from there.

---

Privacy and AI Tools

Major AI tools (ChatGPT, Claude, Gemini) have privacy policies that describe how your conversations are handled. By default, most don't train on your chats, but it's worth checking. Don't share sensitive personal information unnecessarily.

• ChatGPT: conversations aren't used for training by default (opt-in or out in settings).
• Claude: Anthropic does not train on conversations by default.
• Rule of thumb: would you say this to a stranger who might remember it? If not, don't share it with AI.

---

Deepfakes: What to Know

A deepfake is a realistic AI-generated video or audio of a real person, often saying or doing something they never said or did. They are used for fraud, disinformation, and manipulation. Healthy scepticism is your best defence.

• Check the source: deepfakes spread via social media, not official channels.
• Look for micro-inconsistencies: hair, teeth, background blur, unnatural blinking.
• If a video of a public figure says something shocking, search for the original source before believing it.

---

AI and Your Financial Safety

AI is used in investment scams, fake product reviews, and fraudulent customer service impersonation. Legitimate financial institutions use AI — but they don't need your passwords or PIN to help you.

• AI investment scams use 'social proof' (fake reviews, fake testimonials) to build trust.
• Real banks use AI in fraud detection — but never ask for your password.
• Financial Conduct Authority: check the FCA register before investing in anything unusual.

---

Talking to Family About AI Safety

Scam awareness is a family conversation. Share what you know about AI-powered scams with your adult children and grandchildren — and hear their concerns about the scams targeting their generation too.

• Young people are statistically more vulnerable to investment and romance scams than older generations.
• The 'code word' for family emergencies is a useful habit for all generations.
• Action Fraud: 0300 123 2040 — report any suspected fraud immediately.

---

When AI Gets Something Dangerously Wrong

AI can produce dangerous misinformation about medical symptoms, drug interactions, and legal rights. For anything with serious consequences, always verify with a qualified professional.

• Never follow AI advice on medication dosages without pharmacist confirmation.
• Never act on AI legal advice without consulting a solicitor.
• When in doubt: 'What's the authoritative source I should check for this?'