Picture a building that tests its smoke alarms once a year and never runs a fire drill. Everyone ticks a box, the certificate goes in a folder, and nobody knows how the team would react when the alarm goes off.
That’s how many organisations still treat security awareness training. One annual video, one quiz, and one completion report. Meanwhile, the human element was involved in 62% of breaches, according to Verizon’s 2026 Data Breach Investigations Report [1].
Your people aren’t the problem. The training model is. Here’s why it needs a rethink, and what to replace it with.
Why Annual Training Falls Short
Annual training was built for a threat that has moved on. Four things have changed.
- It’s a one-off. Skills fade, and attackers change their lures every month.
- It’s generic. Finance, HR, and the executive team get the same video, even though they face very different attacks.
- It’s email-only. 41% of social engineering breaches in the 2026 DBIR involved vectors other than email [2]. Mobile-based phishing simulations also drew about 40% more engagement than email ones [1]. Voice calls, texts, and collaboration apps are now part of the picture.
- The lures got better. Attackers are using AI to produce more tailored phishing and deepfake-enabled social engineering [3], and KnowBe4 recorded a 17.1% rise in phishing attacks since late 2025 [4].
The result is a workforce that is exposed before anyone has trained them. KnowBe4’s 2026 benchmark puts the global average phish-prone percentage at 33.2% before training, which means about one in three employees is likely to engage with a phishing test [5]. In Australia and New Zealand, it’s about 33% too, and organisations with more than 10,000 employees start at 54.5% [3].
What Continuous Training Looks Like
The fix isn’t more training. It’s a different kind. Think of five shifts.
- Continuous, not annual. Short, regular touchpoints beat one long session. In KnowBe4’s data, the global phish-prone percentage fell from 33.2% to 20.1% within 90 days, and to 4.2% after 12 months [5]. In Australia and New Zealand, it fell to 5.3% [3]. The 90-day mark isn’t the finish line, so the programme can’t stop there [5].
- Tested, not assumed. Simulated phishing shows what people do, not what they say they’d do.
- Personalised, not generic. Training that fits someone’s role and past behaviour sticks. Coaching at the moment of a risky click lands better than a lecture months later.
- Beyond email. Test the channels of attackers use voice, text and collaboration tools.
- Behaviour, not completion. Count what people do, not who watched the video.
One note on the numbers. They come from KnowBe4’s own customer simulations, so treat them as a benchmark rather than a promise. Still, a drop from 33.2% to 4.2% is hard to ignore.
How to Measure a Human Firewall
Completion rates tell you who watched the video. They don’t tell you who would click. Track these instead:
- Phish-prone percentage. The share of employees who engage with a simulated attack [5]. Measure it at baseline, 90 days and 12 months.
- Reporting rate. How many people report a suspicious message, not just avoid it.
- Time to report. How quickly does the first report land after a message arrives.
- Repeat clickers. A small group often drives most of the risk, and it needs targeted support.
- Channel coverage. Whether you’re testing voice, text and collaboration tools, not just email.
The falling click rate is good. A rising report rate is better, because it means your staff are acting as sensors, not just dodging mistakes.
How Mitrais Approaches It
Mitrais helps organisations treat awareness as an ongoing programme, not an annual event. It starts with a baseline: how phish-prone are your people today, and across which teams and channels?
From there, the work is choosing and setting up the right tooling. Platforms such as KnowBe4 bring simulated phishing, training content and real-time coaching together, with reporting on phish-prone percentage over time [6]. The goal is a programme your security team can run month to month, with results the business can see.
And because people are one layer, not the only one, awareness training works best alongside controls like identity protection and zero trust access. That way, one click doesn’t hand the attacker in the keys.
Where to Start
You don’t need to rebuild everything at once. Four steps get you moving.
- Baseline. Run a simulated phishing test to find your starting phish-prone percentage.
- Target. Give high-risk teams and repeat clickers training that fits their role.
- Widen. Add voice, text and collaboration-tool scenarios to your email tests.
- Report on behaviour. Show leadership click rates, reporting rates and time to report each quarter, not completion percentages.
Want to See How Your People Would Hold Up?
If your training is still a once-a-year video, there’s a gap worth closing. Talk to Mitrais about reviewing how your staff are tested and supported today.
Sources
- Help Net Security, Lessons for organizations from the Verizon 2026 Data Breach Investigations Report
- PhishingBox, Verizon 2026 DBIR human risk and phishing awareness
- IT Brief, Australia and New Zealand phishing risk drops after training
- KnowBe4, press release on global phishing susceptibility (Business Wire, July 2026)
- KnowBe4, 2026 Phishing by Industry Benchmarking Report: Findings on Human Risk
- KnowBe4, platform overview