AI-powered cyber threats: how to protect your business

How criminals use AI for phishing, deepfakes, malware and automated attacks, and the practical defences small and mid-sized businesses should put in place.

By the UK Technology Hub teamPublished 4 minute read

The short answer

Criminals use AI to write convincing phishing messages, clone voices and faces, find vulnerabilities faster and automate attacks. Businesses also face attacks on their own AI systems. Defend with strong basics: multi-factor authentication, updates, backups and least-privilege access, plus verification procedures that don't rely on voice or email alone, staff training on AI-enabled scams and secure design of AI tools.

AI makes many tasks easier, including for attackers. Phishing emails no longer have obvious spelling mistakes, voices can be cloned and attacks can be automated at scale. The good news: most defences are proven basics applied consistently.

How attackers use AI

Convincing phishing

AI writes personalised, fluent messages in any language, sometimes using information gathered from websites and social media.

Deepfake voices and video

Criminals impersonate executives or suppliers in calls or video meetings to request payments or access. Our guide to AI voice cloning risks explains this threat.

Faster vulnerability discovery

AI can help attackers find and exploit weaknesses in software more quickly.

Automated attacks

Credential stuffing, fake account creation and scraping at scale.

Synthetic identities

Fake identities combining real and invented information to open accounts.

Attacks on your AI systems

  • Prompt injection manipulating chatbots or agents.
  • Data leakage through AI tools with excessive access.
  • Poisoned content influencing AI outputs.

Secure AI integrations with limited permissions, approvals and monitoring.

Core defences

  1. Multi-factor authentication on all important accounts.
  2. Regular updates for software, websites and devices.
  3. Backups, tested and protected from ransomware.
  4. Least-privilege access for staff, vendors and AI tools.
  5. Email security with authentication, filtering and DMARC.
  6. Endpoint protection on devices.
  7. Monitoring and alerts for unusual activity.

Verification procedures

  • Confirm payment requests and bank detail changes through known, separate channels.
  • Use code words or call-back procedures for sensitive requests.
  • Require dual approval for large payments.
  • Be suspicious of urgency and secrecy.

Staff training

Update training to cover AI-enabled scams: realistic phishing, voice and video impersonation, and fake websites. Encourage staff to verify without fear of embarrassment.

Incident response

Prepare a plan: who to contact, how to isolate systems, how to communicate and how to report incidents to authorities and regulators where required.

Fraud overlaps

Many AI threats aim at financial fraud. Our guide to AI fraud detection covers detection tools and processes.

Using AI for defence

Security tools increasingly use AI to detect anomalies, phishing and malware. They help, but don't replace basic controls and human vigilance.

Small business priorities

If resources are limited, focus first on multi-factor authentication, updates, backups, verification procedures for payments and staff awareness.

A worked example

A design agency received an email that appeared to come from a regular client, asking for a change to the bank account for an upcoming refund. The email was well written and referenced a real project, likely using information from social media and a compromised mailbox. The finance manager followed the agency's rule of verifying bank detail changes by phone using the client's known number. The client confirmed they hadn't sent it. The agency reported the incident, informed the client about the likely compromise and reminded staff of the procedure.

Phishing red flags in the AI era

  • Unexpected requests for money, credentials or data.
  • Pressure to act quickly or secretly.
  • Slight differences in email addresses or domains.
  • Requests to move conversations to unusual channels.
  • Messages that feel right but ask for something unusual.

Good grammar is no longer a reliable signal of legitimacy.

Tabletop exercises

Run short scenario exercises with staff: a deepfake call, a compromised supplier email or a ransomware alert. Practising responses makes real incidents less chaotic.

Vendor and supply chain risk

Attackers target suppliers to reach customers. Ask key vendors about their security, limit their access and monitor shared accounts.

Reporting

Encourage staff to report suspicious messages quickly. Fast reporting helps contain incidents and protect others.

Basic protections still matter

Most attacks, including AI-assisted ones, still exploit basic weaknesses. Multi-factor authentication, prompt software updates, regular backups tested for recovery, least-privilege access and staff awareness stop a large share of incidents. Get these right before investing in advanced tools.

Getting outside help

Small businesses often lack in-house security expertise. Consider a managed IT provider, a security review or government-backed schemes that set out basic controls. Outside help is cheaper than recovering from a serious incident.

How we help

Our maintenance and support team keeps websites and systems updated and monitored, our AI solutions team secures AI integrations and our web app developers build verification and approval workflows into your systems.

Frequently asked questions

How do criminals use AI in cyber attacks?

For convincing phishing, deepfake voices and video, automated attacks and faster vulnerability discovery.

What's the best defence against AI phishing?

Multi-factor authentication, email security, staff training and verifying unusual requests through separate channels.

Can deepfake video calls fool staff?

They can. Verify sensitive requests through known channels, regardless of how convincing a call seems.

This guide is general information, not legal, tax or financial advice, and details such as prices, features and rules can change. Product and company names mentioned are trademarks of their respective owners and are used only to identify those products; UK Technology Hub is not affiliated with or endorsed by them.

Related guides

AI fraud detection for small and online businesses

How AI helps detect payment fraud, account takeovers, fake sign-ups and suspicious transactions, and how smaller businesses can use it without blocking genuine customers.

All AI & automation guides