• Intellery
  • Posts
  • Nigeria launches AI UniPod to drive AI innovation and skills

Nigeria launches AI UniPod to drive AI innovation and skills

PLUS: South Africa bets on decentralised AI in new policy shift

Weekly Roundup

Hi Ayodele here,

This week’s update: Nigeria is creating spaces for innovation through its AI UniPod and improving public engagement with new AI tools, while South Africa is developing a policy approach that balances innovation with oversight. These developments show how countries are laying the groundwork for more practical and responsible AI adoption.

Let us get into it. In this edition:

  • Nigeria launches AI UniPod to drive AI innovation and skills

  • Nigeria launches AI-powered app to improve access to government services

  • South Africa bets on decentralised AI in new policy shift

Around Africa

Nigeria launches AI UniPod to drive ÀI innovation and skills

The Federal Government of Nigeria, in partnership with the United Nations Development Programme (UNDP) and TETFund, officially launched the national rollout of University Innovation Pods (UniPods) this week, with the flagship Artificial Intelligence UniPod at the University of Lagos. Commissioned by Vice President Kashim Shettima, the facility is designed as a "collision space" where students, researchers, and industry leaders can collaborate to move AI ideas from the lab to the marketplace.

This launch marks a decisive move to reposition Nigerian universities as engines of production and enterprise rather than just centers of instruction. The UNILAG hub is the first of a planned national network, with subsequent pods specializing in sectors like Agriculture (Benue), Resilience (Maiduguri), and Manufacturing (Abia) to ensure regional economic strengths are augmented by frontier technology.

Why This Matters

For Nigeria to move from a consumer to a producer of technology, it needs a dedicated pipeline for "sovereign AI" capability. The UniPod initiative creates a structured pathway for research commercialization, ensuring that Nigerian youth aren't just learning about AI but are building scalable, indigenous solutions that address local challenges in finance, health, and governance.

Nigeria launches AI-powered app to improve access to government services

Nigeria has launched CLHEEAN, an AI app meant to help people connect more easily with government services. It lets users ask questions, get policy information, send feedback, report concerns, and join public discussions. The app also includes instant messaging, voice recognition, and support for several local languages. Officials say it is meant to close the communication gap between citizens and public institutions. It was announced by the National Orientation Agency on March 30.

Why it matters

This could make government information easier to reach, especially for people who need faster and more direct support. It also shows how African governments are starting to use AI for public service and civic engagement.

South Africa bets on decentralised AI in new policy shift

South Africa’s new AI policy does not create one big AI regulator. Instead, it gives oversight to existing bodies in areas like finance, health, and data protection. The policy uses a risk-based system, so higher-risk AI tools face stricter checks, while lower-risk tools get lighter rules. It also plans to set up advisory and coordination groups, but these groups will not have direct enforcement power. The goal is to stay flexible, support innovation, and build AI that works better for local languages and local needs. But the article says this approach could lead to uneven enforcement if some regulators are stronger than others.

Why it matters

South Africa is choosing a more flexible path than some other African countries. If it works, it could support innovation while still managing risk, but success will depend on how well different regulators coordinate and enforce the rules.

Around the world

OpenAI Shelves Stargate UK Over Energy Costs

In a significant blow to the UK’s tech ambitions, OpenAI has reportedly put its landmark £31 billion "Stargate UK" investment on hold this week. The project, intended to be a cornerstone of the UK-US AI deal, was stalled due to soaring energy costs and regulatory hurdles. Critics are calling it a wake-up call for governments relying on foreign tech giants to build their sovereign infrastructure.

Anthropic Secures Gigawatt-Scale Compute for 2027

While some projects stall, others are doubling down. Anthropic announced a massive new partnership with Google and Broadcom this week to secure multiple gigawatts of next-generation TPU capacity. With their revenue run-rate reportedly surpassing $30 billion, up from $9 billion just months ago, Anthropic is building a war chest of infrastructure. This deal ensures that future versions of Claude will have the engine room needed to push the boundaries of agentic AI.

Apple Intelligence Guardrails Bypassed in New Attack

Cybersecurity researchers at RSAC successfully hacked Apple Intelligence this week using a combination of "Neural Exec" prompt injections and Unicode manipulation. By tricking the AI into reading malicious instructions backward, they were able to bypass safety filters and access sensitive data. While Apple has already rolled out protections in the latest OS updates, the breach serves as a stark reminder that even privacy-first on-device AI isn't immune to sophisticated adversarial attacks.

Around the Web - Top Picks

  • Project Glasswing & The Mythos Preview

    Anthropic just unveiled Project Glasswing, a massive defensive coalition involving Google, Microsoft, and NVIDIA. It centers on a new, unreleased model called Claude Mythos, which has demonstrated superhuman abilities in finding, and potentially exploiting, vulnerabilities in every major operating system. The project is an urgent attempt to use these offensive capabilities for a global defensive shield before they proliferate to bad actors.

    Read the full article: Project Glasswing & The Mythos Preview

  • Meta’s "Muse Spark" & The Superintelligence Lab
    Meta has officially moved away from the Llama branding for its latest frontier models, launching Muse Spark. This is the first release from their newly formed Meta Superintelligence Lab (MSL), designed to be a multimodal personal superintelligence. Unlike previous models, Muse Spark is built specifically to cite recommendations and content shared across Instagram and Threads, effectively turning your social graph into a reasoning engine.

    Read the full article: Meta’s "Muse Spark" & The Superintelligence Lab

  • The Continuity Breakthrough in AI Filmmaking: 
    A new framework called GIBO WATCH has solved one of the most frustrating problems in AI video: character and scene consistency. By introducing a Context-Retention Engine, the system allows AI to remember character outfits, lighting, and narrative flow across different shots. This marks a significant move toward production-grade AI filmmaking, where the AI acts less like a generator and more like a cinematographer with a memory.

    Read the full article: The Continuity Breakthrough in AI Filmmaking

Seen This?

Anthropic just revealed that its unreleased Claude Mythos model discovered a critical vulnerability in OpenBSD that had remained hidden for 27 years. This "forensic" breakthrough proves that frontier AI can now autonomously identify and weaponize exploits in legacy code that have survived decades of human and automated audits.

Upcoming African AI Events

AI Everything Kenya x GITEX Kenya
Date: May 19–21, 2026
Location: Nairobi, KenyaTast Africa's largest AI and tech event, powered by GITEX, bringing together enterprises, startups, policymakers, and governments with a focus on Kenya's AI-driven economic transformation and digital infrastructure growth.
Link:  AI Everything Kenya x GITEX Kenya  

Deep Learning IndabaX Nigeria 2026
Dates: May 11–14, 2026
Location: University of Ibadan, Nigeria
Theme: Strengthening the Nigerian AI ecosystem through grassroots knowledge sharing. This event focuses on deep learning research, practical workshops on Large Language Models (LLMs), and fostering collaboration among students, researchers, and industry professionals across West Africa.

AI Job Opportunities

  • Yassir - Data Scientist, Lagos Nigeria

  • Nedbank - Data Scientist. Johannesburg

  • Freeland - Data Analyst, Mauritius

  • Mind Detect - ML Researcher, Capetown South Africa

Use Case: How Zapier Central Transforms Team Productivity

In today's fast-paced digital world, managing repetitive tasks across multiple apps is a real challenge. Teams waste valuable time manually moving data, routing requests, and following up across platforms instead of focusing on what actually matters. This is where Zapier Central comes in.

Zapier Central is an AI-powered bot workspace that connects to 6,000+ apps and learns how to handle tasks on your behalf. Think of it as a tireless digital teammate that triages support tickets, drafts responses using live data, and routes work to the right people automatically, across Gmail, Zendesk, Slack, Notion, and more. No more manual handoffs or copy-pasting between tools. Central handles the heavy lifting for you.

Learn how to use here

Reply

or to participate.