Tag Archive for: AI

AI-Powered Cyberattacks Are Here — Here Is How Nigerian Businesses Can Respond

/in , ,

Cybercriminals across the world have picked up a new tool: artificial intelligence. They are using it to write more convincing phishing emails, build malware that changes its code to evade detection, and automate attacks at a scale that was previously impossible. For Nigerian businesses, this is not a distant problem.

Deloitte Nigeria’s Cybersecurity Outlook 2025 states plainly that the “race between AI-powered cyberattacks and AI-driven defence is expected to intensify,” and that Nigerian organisations must treat AI not merely as a defensive tool but as core to their security strategy. The CYFIRMA threat assessment on Nigeria found that in 2025 alone, banking databases, telecom records, and government data were sold on dark web forums. Some listings included over 60 million Nigerian phone records.

A developer in Lagos reviewing a security dashboard

Cybersecurity analysts in Nigeria face a rapidly evolving AI threat landscape. Image: DST

What AI-Powered Attacks Actually Look Like

Polymorphic malware rewrites its own code each time it spreads, making traditional signature-based antivirus tools largely useless. Nigerian IT teams that rely on free or outdated antivirus software face the greatest exposure.

AI-generated phishing now produces emails and SMS messages indistinguishable from authentic bank communications. The days of spotting a scam by its typos are largely over. As ngCERT (Nigeria’s Computer Emergency Response Team) has documented, phishing campaigns within Nigeria’s cyber ecosystem have increased sharply. Read more: Beware the Bengal Cat: A Sinister Search Term.

Deepfakes are being used in audio and video to impersonate executives during financial transactions. A tactic increasingly documented in West African corporate environments.

The Scale of the Problem

Nigeria’s cybersecurity market was valued at $230 million in 2025 and is projected to reach $414 million by 2031, growing at 10.32% annually (Mordor Intelligence, January 2026). The National Information Technology Development Agency (NITDA) estimates Nigeria loses over $500 million annually to cybercrime.

African organisations suffer a significantly higher rate of attacks compared to global averages. Check Point Software’s research shows Africa is frequently used as a testing ground for new attack methods before they are deployed elsewhere.

Practical Steps for Nigerian Businesses

  1. Move beyond passwords. Implement multi-factor authentication (MFA) on all systems, especially for financial and administrative accounts.
  2. Train your employees. IBM research consistently shows that over 95% of cyberattacks involve human error. Related: New Employees Can Be a Cybersecurity Risk.
  3. Move to cloud-native security. Cloud-delivered security controls now represent 57.2% of cybersecurity spending in Nigeria, growing at 20.4% annually.
  4. Implement zero-trust architecture. Zero trust means no user or device is trusted by default — access is verified continuously.
  5. Choose your hosting carefully. A host with DDoS protection, automatic backups, and a Web Application Firewall reduces your attack surface. Read: Don’t Let Bad Hosting Sink Your Website.

Where to Report Incidents

  • ngCERT (cert.gov.ng) — Nigeria’s national response team
  • NITDA (nitda.gov.ng) — for data breaches under the Nigeria Data Protection Act
  • EFCC — for financial fraud related to the breach

Stair-scaling robot dog to be sold in 2019

/in

The yellow robot dogs that so unsettled people in February by opening a door for one another can now easily walk up and down stairs on their own.

Boston Dynamics CEO Marc Raibert — who founded the robot company known for building machines that mimic real animal and human movements — said Friday at the TechCrunch Robotics Sessions conference at UC Berkeley that the robots, called SpotMini, will be ready for purchase in 2019.

While declining to offer a price on the 66-pound, relatively quiet, and uniquely agile robot dogs, Raibert did say that they’re designed to be deployed in commercial office building-like settings — and then eventually homes.

Raibert presented the above video at the conference, which showed the dogs navigating the Boston Dynamics lab on their own. But, although the robot is traversing autonomously, this shouldn’t be confused with “AI,” the hot, and often misapplied buzzword in today’s tech realm.

To be clear, these robots have been given sophisticated motion detection software that maps out the boundaries of where they can go and where they absolutely shouldn’t go. That’s why the robots so confidently scaled the stairs, while staying clear of the edges. They “see” a navigational map in real time, which appears in the video at the bottom left of the screen.

“It knows it can’t bump into the railings,” said Raibert.

Today, the term “AI” is most accurately used to describe software — like image recognition — that is fed a wealth of information, which it can then use to make well-informed predictions on its own. This often involves being wrong, and self-correcting.

For example, in 2017 Google’s DeepMind researchers built an AI program that, over the course of a month, taught itself to become the world’s most dominant Go (an ancient Chinese board game) player, without any previous knowledge of how to win.

Some of today’s leading AI researchers, however, will make the argument that this isn’t true intelligence, but rather impressively complex machines performing sophisticated number-crunching.

Atop the robot in the video is also a small, white rectangular box, which shows how any (capable) third-party users can make their own hardware devices — like an enhanced radio receiver or specialized camera — and attach it to the robot for their specific surveillance needs.

The constant movement of four legs, however, drains the robots’ batteries pretty quickly. They’re capable of moving around for 90 minutes before needing to recharge.

Mashable