42AI (Pty) Ltd · South Africa

Turn human cyber risk into measurable learning.

42AI is a South African AI, cybersecurity and technology innovation company. Our first product is planned to help organisations simulate phishing, measure how people respond, and teach what to do next. We are preparing to build the MVP.

Pre-MVP · interactive conceptSeeking future pilot partners
42AIAWARENESS   /   SIMULATION   /   STRONGER HUMANS
MESSAGE INSPECTOR · sample messagetry it
From HR Services <payroll@company-hr-update.invalid>
Subject Review your payroll details by 17:00

Hi team, a payroll update must be acknowledged today. Staff who do not respond may see a delay in this month's payment.

Tap the link text to see where it really goes, then mark what looks wrong.

Fictional training example. Responses are recorded only in this page’s temporary demo state. Nothing is sent; Reset or reload clears them.

The company

One company, two business lines.

42AI plans two business lines: products and professional services. Products give organisations repeatable tools. Services turn real customer problems into insight, trust and better products.

Products

Planned repeatable tools for cybersecurity awareness, human-risk learning and conversational AI.

Professional Services

Planned advisory, development, support and training shaped around customer needs.

Each pillar is labelled with its real stage. We build the cybersecurity product first and grow the other pillars in phases, as capacity and customer evidence allow.

The first product

A continuous human-risk learning loop.

One-off awareness training is hard to measure. The planned product follows a cycle that repeats. This walkthrough illustrates intended behaviour; simulation, measurement and lessons are not live.

Open separate cybersecurity demoConcept demonstration · fictional data · Dashboard, Campaigns, Human Risk Profiler, Micro-Learning, Reports and Compliance.

Built around South African working life.

Planned training examples include a fake SARS refund, a courier delivery-fee SMS, a bank verification request, and a supplier invoice with changed banking details. These are scenario ideas, not partnerships or endorsements.

Pause before acting. Verify requests through an independently known number or official app. At work, follow your organisation’s reporting process. Never share a password or one-time PIN in response to a message.

Governance & trust

Compliance alignment.

Design references for our planned products and professional services.

ISO/IEC 27001

Information security management: risk assessment, awareness, access control and continual improvement guide our intended approach.

ISO/IEC 42001

AI management: accountability, responsible use, human oversight and AI risk review inform planned AI-enabled work.

ISO/IEC 20000-1

Service management: defined service responsibilities, incident handling and service improvement inform planned delivery practices.

Design alignment, not certification. These standards are design references only. 42AI holds no certification or accreditation and is not a certification body. This website does not demonstrate conformity or guarantee customer compliance.

How we work

Responsible by design, honest by default.

These principles guide the planned MVP. They describe our intended approach, not verified product capabilities.

Learning, not blame

Simulation results lead to targeted education and a clear way to report, not to punishment.

Privacy by design

Collect only what a pilot needs, tell staff what is happening, restrict access and agree how data is deleted.

Evidence you can act on

Reports show response and learning trends for management, with measures agreed with each organisation first.

Our claims, stated plainly

Is 42AI live?

Not yet. We are at pre-MVP concept stage. The walkthrough uses pre-written scenarios and fictional data.

Are you certified?

No. We use ISO/IEC 27001, 42001 and 20000-1 as design references and hold no certification or accreditation. We are not a certification body.

Where will data be hosted?

Hosting and AI-processing arrangements are being evaluated. We design with South African data-protection requirements in mind and make no hosting or compliance guarantee yet.

Does AI run on this website?

No. AI-generated scenarios are the intended product direction and still need to be built and tested. The inspector runs locally with fixed examples and does not contact an AI model or send its results. Contact links open your email app. Your hosting provider may process standard access logs; Google Fonts receives requests when fonts load. The theme preference is stored only in your browser.

Roadmap

Build the focus first, then widen.

Three phases, each with a gate. Nothing moves forward until its gate is met.

Phase 1

Validate

  • Customer discovery conversations
  • Product specification, architecture and threat model
  • Secure MVP of the cybersecurity product
GateThreat model reviewed and MVP hardened before any pilot.
Phase 2

Prove

  • Small, controlled pilots with agreed measures
  • First scoped advisory engagements
  • Honest results, including what did not work
GateWritten success measures and data-handling terms agreed with each pilot organisation.
Phase 3

Expand

  • Broaden the product from pilot evidence
  • Mature Governance & Trust services
  • Explore Conversational AI and Innovation
GatePilot evidence and a secure foundation.
Partner with us

Help shape what we build.

42AI is founder-led and early. The most useful help is specific. Choose how you might get involved.

Start a conversation

+27 79 295 3578

Opens your email app with a draft. Review and send it there. If no email app opens, use the address below. Please do not include passwords or sensitive incident details.