The Minister of Communications, Innovation and Digital Economy, Dr Bosun Tijani, disclosed this during the Crisis Management Advocacy Month 2026 conference in Lagos, confirming that the policy framework has moved beyond the draft stage.

According to him, the strategy will provide the legal and regulatory backing needed to enable Nigeria to become competitive in the global AI space, while ensuring innovation is balanced with accountability.

“I am pleased to share that Nigeria now has a National Artificial Intelligence Strategy. It is no longer a draft; it is a fully developed strategy,” the minister said.

Local Language AI Takes Centre Stage

A major highlight of the strategy is the development of a government-backed Large Language Model (LLM) tailored to Nigeria’s linguistic landscape.

Unlike many global AI systems, the proposed model is being designed to understand and communicate in indigenous languages such as Hausa, Igbo and Yoruba, alongside English.

The development is aimed at closing the digital gap for millions of Nigerians who are often excluded from emerging technologies due to language barriers.

In relying on local datasets, the government hopes to build AI tools that better reflect the country’s cultural and social realities.

AI to Drive Crisis Prediction and Response

Beyond communication, the government is positioning AI as a tool for national preparedness and crisis management.

Tijani noted that advanced AI systems will be deployed to analyse large volumes of data, identify patterns and predict potential risks before they escalate.

“Crises are inevitable. What separates those who succeed is preparedness,” he said.

To support these capabilities, the Federal Government is rolling out an ambitious 90,000-kilometre fibre optic network to improve connectivity nationwide and enable real-time data processing.

Issues Over AI-Driven Threats

Despite the positivity, industry stakeholders have warned about the growing risks associated with artificial intelligence.

Group Managing Director of CMC Connect LLP, Yomi Badejo-Okusanya, said organisations are increasingly vulnerable to AI-powered disruptions, including deepfakes, cloned voices and manipulated content.

“Today, crises are faster than facts and louder than truth. The real crisis is the widening gap between how fast threats evolve and how slowly organisations respond,” he said.

He added that more than 60% of organisations have already experienced AI-related crises, while nearly 70% of media institutions remain ill-equipped to respond effectively.

A Defining Moment for Nigeria’s Tech Ecosystem

For Nigeria’s growing digital economy, the National AI Strategy is expected to serve as a roadmap for investment, governance and ethical standards.

Approval by the National Assembly will not only formalise the framework but also signal Nigeria’s readiness to build and deploy indigenous AI systems at scale.

However, experts say the success of the initiative will depend largely on public trust, transparency and the country’s ability to manage emerging risks tied to artificial intelligence.

The commitment was disclosed during a high-level meeting in Abuja between executives of MTN Group and Nigeria’s Minister of Communications, Innovation and Digital Economy, Dr. Bosun Tijani.

The MTN delegation was led by Group President and CEO Ralph Mupita, alongside MTN Nigeria CEO Karl Toriola and Chief Corporate Services & Sustainability Officer Tobe Okigbo.

While the telco did not disclose the size of the planned investment, executives signalled that the new funding cycle will prioritise network expansion, improved service quality, and digital infrastructure upgrades, critical enablers for economic productivity in sectors such as fintech, e-commerce, manufacturing, and logistics.

Why infrastructure matters for a $1 trillion economy

Nigeria’s digital economy currently contributes roughly 18–20% to GDP, according to government estimates.

But scaling that contribution significantly will require more reliable broadband penetration, lower latency networks, and improved rural connectivity.

Speaking at the engagement, Mupita noted that Nigeria’s macroeconomic ambition cannot be achieved without a communications backbone capable of supporting enterprise digitisation and industrial growth.

“For Nigeria to achieve its economic targets, the digital infrastructure must be robust enough to power innovation at scale,” he said, framing telecom infrastructure as foundational rather than complementary.

The timing is strategic. Over the past year, regulators approved spectrum trades and implemented tariff adjustments that operators had long argued were necessary to sustain heavy capital investments amid inflation and currency volatility.

Mupita described these reforms as instrumental in stabilising the industry and enabling long-term planning.

Nigeria remains MTN Group’s largest and most strategic market, contributing a significant share of group revenue. A stable regulatory climate makes it easier for the operator to justify deeper capital commitments.

Government support, but with conditions

Minister Tijani welcomed the pledge but struck a note of accountability. Government backing, he said, will remain tied to measurable service improvements and consumer protection standards.

The ministry has walked a delicate line in recent months, balancing operator sustainability with affordability concerns and maintaining competition in a sector where consolidation risks could reduce consumer choice.

Tijani reiterated that Nigeria’s telecom future must remain competitive, cautioning against market dynamics that tilt toward monopolistic dominance.

“A diversified industry is essential for innovation and fair pricing,” he said.

AI, LLMs, and the next telecom frontier

Beyond infrastructure, discussions extended to artificial intelligence and Nigeria’s ambition to play a stronger role in Africa’s AI ecosystem.

Tijani, recently recognised internationally for his work in digital policy, challenged MTN to evolve beyond traditional connectivity and actively participate in shaping Africa’s AI future.

Central to this conversation is Nigeria’s National Large Language Model (LLM) initiative, which aims to build AI systems reflective of local languages and contexts.

Mupita signalled MTN’s interest in deeper collaboration around AI integration within telecom services, noting that artificial intelligence is rapidly becoming central to network optimisation, fraud detection, customer service automation, and enterprise solutions.

For telecom operators globally, AI is no longer experimental, it is operational infrastructure.

From cables to capabilities

The meeting concluded with both parties emphasising digital skills development as a parallel priority. As MTN advances its “Ambition 2030” strategy, the company pledged stronger collaboration with the ministry on workforce readiness.

The subtext is clear: expanding fibre networks and 5G coverage is insufficient without engineers, data scientists, cybersecurity specialists, and AI researchers to manage and monetise that infrastructure.

For Nigeria’s $1 trillion economy goal, telecom infrastructure is only one pillar.

The others, policy coherence, human capital development, competitive markets, and sustained foreign investment, will determine whether ambitious projections translate into measurable growth.

MTN’s renewed capital commitment signals confidence. The question now is whether execution, regulation, and macroeconomic stability can keep pace.

These are required to develop the continent’s own large language models (LLM) to power AI-driven solutions for Africa’s 1.5 billion people, who otherwise risk being sidelined by the global AI ecosystem.

Speaking on ‘The Y’ello Chair Vodcast: Your link to the African continent’, Dr Bosun Tijani, Nigeria’s minister of Communications, Innovation and Digital Economy said that to leapfrog AI in Africa, a collaborative public/private effort was urgently needed to fund the academic research into the continent’s many languages.

He challenged MTN Group, which has operations in 16 markets, 15 of them in Africa, to mobilise resources for this.

“We like these kinds of partnerships. Challenge accepted,” said Ralph Mupita, MTN Group President and CEO in the vodcast, which was filmed on the sidelines of the 80th United Nations General Assembly in New York.

It was hosted by Angela Wamola, who is the head of sub-Saharan Africa for the GSMA mobile industry association.

The vodcast followed the launch of the Nigerian Atlas for Languages & AI at Scale (N-ATLAS). This is an open-source multilingual LLM designed to understand and generate Nigeria’s diverse voices, digitising and preserving the country’s linguistic richness and creating datasets for AI solutions.

More than 500 languages are spoken in Nigeria, the continent’s most populous country.

N-ATLAS is a public/private initiative of the government of Nigeria and Awarri Technologies.

The ATLAS framework is open and available to other African countries, providing a platform for innovation in local languages and in so doing, transform education, health, commerce and governance.

“We have to avoid the risk of Africans being a digital underclass,” Mupita said of the work to ensure that citizens didn’t feel excluded on the continent where there are more than 2 000 languages and most are poorly represented in the global AI ecosystem.

He said the digital economy was the “best bet” to ensure that citizens have dignity, hope and opportunity.

“The outcomes we want are that people are digitally included, economically included and that they have dignity. This dignity point for me is very important because poverty can include all sorts of indignity, but embracing technology should take all that away,” he said.

The Y’ello Chair Vodcast is a platform where leaders and changemakers share insights, challenge norms, and drive Africa’s digital future. For episode 2, see here.

Key Facts

• 16 markets: MTN Group operates in 16 markets, 15 of them in Africa.
• 500+ languages in Nigeria: Nigeria is Africa’s most linguistically diverse country.
• 2,000+ languages in Africa: Most are poorly represented in AI models.
• 1.5 billion people: Africa’s population risks exclusion from the benefits of AI.
• N-ATLAS: Joint initiative between the Nigerian government and Awarri Technologies to digitise and preserve Nigerian languages.

The previously unknown malware is delivered via a phishing site pretending to be the official DeepSeek homepage that is promoted via Google Ads.

The goal of the attacks is to install BrowserVenom, a malware that configures web browsers on the victim’s device to channel web traffic through the attackers servers, thus allowing to collect user data – credentials and other sensitive information.

Multiple infections have been detected in Brazil, Cuba, Mexico, India, Nepal, South Africa and Egypt.

DeepSeek-R1 is one of the most popular LLMs right now, and Kaspersky has previously reported attacks with malware mimicking it to attract victims.

DeepSeek can also be run offline on PCs using tools like Ollama or LM Studio, and attackers used this in their campaign.

Users were directed to a phishing site mimicking the address of the original DeepSeek platform via Google Ads, with the link showing up in the ad when a user searched for “deepseek r1”. Once the user reached the fake DeepSeek site, a check was performed to identify the victim’s operating system.

If it was Windows, the user was presented with a button to download the tools for working with the LLM offline. Other operating systems were not targeted at the time of research.

After clicking on the button and passing the CAPTCHA test, a malicious installer file was downloaded and the user was presented with options to download and install Ollama or LM Studio.

If either option was chosen, along with legitimate Ollama or LM Studio installers, malware got installed in the system bypassing Windows Defender’s protection with a special algorithm.

This procedure also required administrator privileges for the user profile on Windows; if the user profile on Windows did not have these privileges, the infection would not take place.

After the malware was installed, it configured all web browsers in the system to forcefully use a proxy controlled by the attackers, enabling them to spy on sensitive browsing data and monitor the victim’s browsing activity.

Because of its enforcing nature and malicious intent, Kaspersky researchers have dubbed this malware BrowserVenom.

“While running large language models offline offers privacy benefits and reduces reliance on cloud services, it can also come with substantial risks if proper precautions aren’t taken. Cybercriminals are increasingly exploiting the popularity of open-source AI tools by distributing malicious packages and fake installers that can covertly install keyloggers, cryptominers, or infostealers. These fake tools compromise a user’s sensitive data and pose a threat, particularly when users have downloaded them from unverified sources,” comments Lisandro Ubiedo, security researcher with Kaspersky’s Global Research & Analysis Team.

To avoid such threats, Kaspersky recommends:

• Check the addresses of the websites to verify that they are genuine and to avoid a scam.
• Download offline LLM tools only from official sources (e.g., ollama.com, lmstudio.ai).

• Avoid using Windows on a profile with admin privileges.
• Use trusted cyber security solutions to prevent malicious files from launching.

Consistent with the company’s commitment to open-source AI, the Granite models are released under the permissive Apache 2.0 license, making them unique in the combination of performance, flexibility and autonomy they provide to enterprise clients and the community at large.

The Granite 3.0 release reaffirms IBM’s commitment to building transparency, safety, and trust in AI products.

The Granite 3.0 technical report and responsible use guide provide a description of the datasets used to train these models, details of the filtering, cleansing, and curation steps applied, along with comprehensive results of model performance across major academic and enterprise benchmarks.

Raising the bar: Granite 3.0 benchmarks 

The Granite 3.0 language models also demonstrate promising results on raw performance.

The Granite 3.0 models were trained on over 12 trillion tokens on data taken from 12 different natural languages and 116 different programming languages, using a novel two-stage training method, leveraging results from several thousand experiments designed to optimize data quality, data selection, and training parameters.

IBM is also announcing an updated release of its pre-trained Granite Time Series models, the first versions of which were released earlier this year.

These new models are trained on 3 times more data and deliver strong performance on major time series benchmarks.

Introducing Granite Guardian 3.0: ushering the next era of responsible AI  

IBM is also introducing a new family of Granite Guardian models that permit application developers to implement safety guardrails by checking user prompts and LLM responses for a variety of risks.

The Granite Guardian 3.0 8B and 2B models provide the most comprehensive set of risk and harm detection capabilities available in the market today.

In addition to harm dimensions such as social bias, hate, toxicity, profanity, violence, jailbreaking and more, these models also provide a range of unique RAG-specific checks such as groundedness, context relevance, and answer relevance.

Assistants to Agents: realizing the future for enterprise AI 

IBM is advancing enterprise AI through a spectrum of technologies – from models and assistants, to the tools needed to tune and deploy AI specifically for companies’ unique data and use-cases.

IBM is also paving the way for future AI agents that can self-direct, reflect, and perform complex tasks in dynamic business environments.

IBM continues to evolve its portfolio of AI assistant technologies – from watsonx Orchestrate to help companies build their own assistants via low-code tooling and automation, to a wide set of pre-built assistants for specific tasks and domains such as customer service, human resources, sales, and marketing.

Today IBM also unveiled the upcoming release of the next generation of watsonx Code Assistant, powered by Granite code models, to offer general-purpose coding assistance across languages like C, C++, Go, Java, and Python, with advanced application modernization capabilities for Enterprise Java Applications.

Expanded AI-powered delivery platform to supercharge IBM consultants with AI

IBM is also announcing a major expansion of its AI-powered delivery platform, IBM Consulting Advantage.

The multi-model platform contains AI agents, applications, and methods like repeatable frameworks that can empower 160,000 IBM consultants to deliver better and faster client value at a lower cost.

As part of the expansion, Granite 3.0 language models will become the default model in Consulting Advantage.

Another key part of the expansion is the introduction of IBM Consulting Advantage for Cloud Transformation and Management and IBM Consulting Advantage for Business Operations.

Each includes domain-specific AI agents, applications, and methods infused with IBM’s best practices so IBM consultants can help accelerate client cloud and AI transformations in tasks, like code modernization and quality engineering, or transform and execute operations across domains, like finance, HR and procurement.

In today’s digital landscape they struggle with islands of data spread across various systems, leading many workers to not trust the data used to train AI systems and experience difficulty to get what they want out of them.

According to Salesforce research, only 28% of applications are connected, and over 80% of business leaders struggle with data fragmentation and data silos.

While three-quarters of workers surveyed in the recent “Your Data, Your AI” survey from Salesforce believe accurate, complete, and secure data is critical to building trust in AI, more than half do not trust the data used to train AI systems today.

And nearly 60% of AI users worldwide find it difficult to get what they want out of AI, the report found.

The future of enterprise AI isn’t about more data – it’s about the right data. When AI is grounded in a company’s own data, it delivers more useful results and ultimately drives greater trust and adoption.

Only by consolidating their data will companies be able to fully understand the complete customer journey. A trusted data foundation and integrating AI into workflows across the enterprise are key ingredients needed for AI success.

Deploying these together, companies can unlock enterprise deployments at scale and drive measurable outcomes from AI automation, personalisation, and performance optimisation, including higher sales productivity, faster customer service resolutions, higher-conversion marketing campaigns.

Building a trusted data foundation

For AI to live up to the hype, large language models (LLMs) must be grounded in trusted enterprise data. However, with data trapped in disconnected silos, wholesale digital transformation and value realisation remains elusive.

Prospects are worse when the data being used to ground AI models is incomplete, incorrect, or irrelevant — leading to inconsistent, incorrect results.

Unlocking the power of trapped data enables better analysis, decision-making, and AI automation, grounding customer and business data and metadata — a common language that integrates all applications — in ways that deliver trusted, outcome-oriented results without expensive model training.

Take, for example, real-time data that a prospective customer has just visited a company’s website. Previously, sales reps would have had no way of knowing this without manually pulling data into a custom report.

Real-time data brings actionable insights, allowing for immediate customer engagement, resulting in higher conversion rates, revenue growth, and customer satisfaction.

Trust is a key component of successful enterprise AI deployments. By unifying and cleansing their data, companies can ensure that AI models operate on the most accurate information.

At Salesforce, we have engineered trust into every Salesforce application through our Einstein Trust Layer, a core part of the Einstein 1 Platform.

The Einstein Trust Layer includes data masking to ensure data privacy protection, a zero-retention architecture to ensure data is never learned by AI models or stored outside Salesforce, an LLM audit trail, and keeps humans at the helm of every AI interaction.

We have also built-in a feedback loop that continuously improves model accuracy and relevance, and this feedback data is automatically logged in Data Cloud.

Integrating AI into the flow of work

The need to deliver AI in the flow of where companies’ sales, service, marketing, commerce, developer, and other employees work explains why they’re leaning into conversational assistants, for their employees to interact with any data or workflow across their enterprise.

With specific customer data, employees can generate useful responses which are automatically grounded in all of their organisation’s trusted data and metadata.

From generating customer campaigns, to answering service questions, everything is personalised, based on consolidated data – all securely within the confines of their company’s data and business processes.

The powerful combination of data and CRM makes these personalised customer experiences possible. For today’s consumer, milliseconds matter.

The cost of not keeping up with them could be lost sales opportunities, poor social media reviews, or a disconnect in healthcare delivery.

While generative AI is still in its early stages for most companies, the potential for true enterprise transformation is immense. Those that can put in a foundation of data and trust, and offer AI in the flow of where their employees work, will be able to shift from pilot to production and realise tremendous value, employee satisfaction, customer loyalty, and business growth.

[Featured Image Credit]

With over 100 meticulously designed metrics, it is the most comprehensive platform that allows developers and organizations to evaluate and compare LLMs effectively and establish essential guardrails for LLMs and Retrieval Augmented Generation (RAG) applications. 

These tests assess various aspects including Relevance & Understanding, Content Quality, Hallucination, Safety & Bias, Context Relevance, Guardrails, Vulnerability scanning, along with a suite of Metric-Based Tests for quantitative analysis.

The RagaAI LLM Hub is uniquely designed to help teams identify issues and fix them throughout the LLM lifecycle, be it a proof-of-concept or an application in production.

From understanding the quality of datasets to prompt templating and the choice of LLM architecture or vectorDBs, RagaAI LLM Hub identifies issues across the entire RAG pipeline. 

This is pivotal for understanding the root cause of failures within an LLM application and addressing them at their source, revolutionizing the approach to ensuring reliability and trustworthiness. 

“At RagaAI, our mission is to empower developers and enterprises with the tools they need to build robust and responsible LLMs,” said Gaurav Agarwal, Founder and CEO of RagaAI. 

“With our comprehensive open-source evaluation suite, we believe in democratizing AI innovation. Together with the enterprise ready version, we’ve created a comprehensive solution to enable organizations to navigate the complexities of LLMs deployment with confidence. This is a game-changing solution that provides unparalleled insights into the reliability and trustworthiness of LLMs and RAG applications.”

The RagaAI LLMs Hub is already utilized across industries like E-commerce, Finance, Marketing, Legal, and Healthcare, and the platform supports developers and enterprises in various LLM applications including chatbots, content creation, text summarization, and source code generation. 

For instance, one customer came to RagaAI after encountering hallucinations and incorrect outputs whilst developing a customer service chatbot. Leveraging RagaAI LLM Hub’s comprehensive metrics, they pinpointed and quantified the hallucination issue in the RAG pipeline.

Additionally, the platform adeptly identifies nuanced customer issues and recommends specific areas within the pipeline for resolution.

The RagaAI LLM Hub helps in setting guardrails, ensuring data privacy and legal compliance, including transparency regulations and anti-discrimination laws. It plays a vital role in promoting ethical & Responsible AI practices, such as in sensitive sectors like finance, healthcare, and law. Additionally, it helps mitigate reputational risks by adhering to societal norms and values.

The launch comes following RagaAI’s successful funding round in January 2024, underscoring the company’s momentum and commitment to advancing LLMs quality assurance.

The universal challenge has shifted from “Can we build it?” to “Can we trust it?”

The established playbook for managing AI, known as MLOps, is proving insufficient for this new wave of technology.

We are witnessing a rapid shift towards a new discipline, often called LLM Ops, as new tools emerge daily to handle the unique challenges of language models.

Unlike traditional AI, these models are unpredictable, and our old methods can’t keep up. At the same time, a new security flaw called prompt injection has left these powerful systems dangerously exposed. We are building glass houses without locks, and the consequences of this fragility will be felt worldwide.

This is why the conversation must evolve. The defining feature of a successful AI product in 2024 will not be its power, but its predictability and safety.

This pivot from a model-centric world to a systems-centric one gives rise to a critical new discipline: AI Assurance.

This is not just a technical challenge. It is a global imperative. The solutions we engineer for AI assurance will have worldwide implications, shaping how industries from finance to healthcare deploy this technology responsibly.

There is no better place to spearhead this effort than the UK. With its proactive regulatory landscape, the nation is fostering the exact environment of innovation and responsibility needed to solve these issues.

This is precisely why I moved to the UK to pursue my Master’s degree. It was a strategic decision to position myself at the heart of the solution.

It allows me to learn from the world’s leading experts and dedicate my thesis to tackling the core challenges of building reliable and secure large language models (LLMs).

The future is not about building bigger models. It’s about building them right, and I came here to be at the forefront of that work.