Anthropic shared a report on detecting and countering AI misuse that covers AI misuse events from December 2025 to August 2026. The report shows how AI has been used for purposes that can be deemed illegal or even harmful to governments and people.
Anthropic is an AI safety and research company that describes itself as investing in building reliable, interpretable, and steerable AI systems, including the well-known Claude.ai. The purpose of the unveiled Anthropic report was to show how AI misuse occurred across seven harm areas: cyber operations, influence operations, surveillance, scams and fraud, biological misuse, conventional weapons development, and distillation in different parts of the world.
I’d urge those interested to read and see how the development in the information era, driven by Artificial intelligence, has moved forward and how fast that is. Nevertheless, I’d like us to ask ourselves two questions we may have chosen not to ask ourselves amid the excitement of technological development: Is Tanzania well prepared in this AI era? How is Tanzania positioned to detect, combat, or regulate AI use in day-to-day activities?
Tanzania in the AI space
In Tanzania, AI is well known, and it is safe to say it is being used differently in terms of how and to what extent. LLMs are the most-used AI systems, at least at the personal level, with products from Chinese- and USA-based companies. These include ChatGPT, Gemini, Claude, Kimi, DeepSeek and others.
Mostly, these systems are used for text generation, text analysis, chatbots and virtual assistants, as well as text summarisation and grammar checking. However, AI use goes beyond that. The International Association of Business Analytics Certification explains that AI has branched further and covers more than how AI has been used so far.
These branches include natural language processing, machine learning, computer vision, robotics and expert systems. My understanding is that LLMs fall in the NLP branch, which involves teaching computers to understand and interact with human languages.
READ MORE: A Genuine Lever or More of the Same? Tanzania’s AI Education Guidelines at a Crossroads
So far, Tanzania has done so little to position itself as a country that embraces development in the artificial space within the industry. What we have seen as development in Tanzania has been academics pushing research studies, research projects, and funded academic projects that create impact on academic grounds and ignite that spark about AI, without the possibility of further development and adoption in industry.
I’m confident in saying the industry side of our research and development has been dormant in embracing the ongoing global developments in the AI space due to a lack of experts, or at least non-academic think tanks, in the AI space.
Maybe this lag is because, as a country, we were still behind in research and innovation, have few data scientists, and because a department like this is rarely present in many of the known big companies in the country.
Take a good example: the most-used LLM models/systems. These are not products of academic institutions, even though universities have not lagged in the countries where these systems were developed. Universities have focused more on the efficiency and efficacy of these models, pushing for innovation and improvement, but they are not the champions of this development.
This balance between academia and industry in the countries where AI has been embraced shows a systemic balance that Tanzania lacks, or even so more African countries don’t have, creating a vacuum that will genuinely come and bite back because of the lack of this awareness and action-taking. However, I’d be less appreciative if I didn’t acknowledge what academics have done.
So far, universities have linked with projects that developed chatbots, patented AI-related discoveries, and developed methods for using AI in health service provision. This development has even led the military to view universities as an important technology collaborator.
Nevertheless, companies have integrated chatbots to support service delivery, and the government has repeatedly emphasised the need to incorporate AI into its daily activities and intended service provision to people. Most recently, the government bid farewell to 44 Tanzanian students travelling to Ireland and Russia to pursue specialised studies in Artificial Intelligence (AI) and Data Science under the national scholarship programme.
Remaining vacuum
Yet, a vacuum remains, and this vacuum is why it is imperative to be concerned about both the pace of global change and how AI is being used. The biggest reason for this vacuum has been the lack of policy to define and guide how AI can be used in government and in general use in Tanzania by students, companies, and other actors.
I’ll note that developing a policy that covers all sectors could be difficult because AI use differs, which may explain why it was easier to develop guidelines in education and the health sector on how AI can be used in these two sectors.
Perhaps other important sectors, such as finance and service provision across government and private entities, would have followed, leveraging the presence of AI tools and AI development. Sadly, that hasn’t happened, and this Anthropic report shows how unsafe we can be without such efforts.
To illustrate the risks of unpreparedness, the Anthropic report details how AI has been misused against Africa, with cases in Kenya and several other African countries, primarily in the context of influence operations (astroturfing, fake news, propaganda) and surveillance.
Kenya’s case is notable for being a domestic operation, whereas CAR and DRC were targeted by foreign or commercial actors. These examples illustrate the diverse and evolving landscape of AI-enabled threats across Africa, where the IMF explicitly categorises many African countries as importers of AI technology.
READ MORE: Tanzania Court Adopts Artificial Intelligence (AI) in its Processes
The IMF World Economic Outlook report further stresses this categorisation, showing that African countries stand on the wrong side of major global economic shocks driven by AI-enabled demand and technology supply chains.
Next victim?
Tanzania, falling in this category of poor African countries, could be the next victim of such activities, including surveillance, propaganda and fake news, or even cyberattacks, given the country’s current state and weak institutional preparedness for AI use and regulation.
Understanding these risks and attacks may not be effective without an AI policy that clearly guides how AI use can be regulated and the laws that direct how misuse can be combated to ensure safe use and protection from attacks and misuse.
Perhaps Tanzania could learn from existing regulations, as it did with the Data Protection Act, which appears to be based on EU personal data protection regulations. Even so, Kenya could be a closer model, with its 2025-2030 AI strategy. Regional thinking around the East African Bloc as a collective effort may be an easier route.
Another challenge for Tanzania in the AI space is the competition between two leading AI countries: China and the USA. The USA is home to Microsoft (Copilot), Anthropic (Claude.ai), Google (Gemini), X (formerly Twitter) (Grok), and OpenAI (ChatGPT).
China, by contrast, is home to DeepSeek (DeepSeek), Moonshot AI (Kimi), Alibaba Group (Qwen), ByteDance (Doubao), Baidu (Ernie), and Zhipu (Zhipu AI). These companies make these two countries the leading competitors in the AI space, despite other countries that have established themselves in this competition, including the United Kingdom, Japan, South Korea, Russia, Canada, India, Turkiye, Israel, and Germany.
READ MORE: From Scaling Up to Integrating Tech-Centric Interventions in Tanzania’s Schools
The interesting part of this competition is the evidence that companies based on the continent providing AI services remain minimal, even though the AI ecosystem involves critical components that are abundantly available on the continent.
In retrospect, Nigeria, South Africa, and Kenya lead the continent as Africa’s dominant AI hubs, together accounting for over 60 per cent of all tracked startups. Beyond the continent’s position, another important factor is the way competition between China and the USA is trying to form alliances that could divide the continent into two blocks, as in the early days of the Cold War.
China has established a global AI cooperation body that aims to promote international cooperation and global governance in artificial intelligence. As of July 2026, 29 countries were believed to have signed an agreement to establish the World Artificial Intelligence Cooperation Organisation.
The USA, meanwhile, launched Pax Silica, an international initiative aimed at securing global supply chains for semiconductors, artificial intelligence (AI) infrastructure, and critical minerals while reducing reliance on China.
In the initiative, the USA works with partner nations to build a secure, prosperous, and innovation-driven silicon supply chain. These two positions held by China and the USA present a compelling diplomatic conundrum that Tanzania might find itself in the global political economy of AI.
One country in a similarly complex diplomatic position is Kazakhstan, which has signed a treaty with China and is a partner with the USA in the Pax Silica. This multi-vector involvement has reportedly put it at odds with the US, which requires its Pax Silica partners to choose between China or the US, not both, as AI partners, a situation that has been rebuffed.
READ MORE: NECTA Cancels Results of 58 Students Caught Using ChatGPT in Exams
This ultimatum originates from the competitive environment between these two leading AI powers and the fact that AI partnerships are becoming a form of influence in other economic and diplomatic arenas in the global political economy of AI, forcing AI technology-importing countries into a difficult position.
A different challenge
During the Cold War, Tanzania was a non-aligned country through the Non-Aligned Movement, but this AI partnership situation presents a different challenge that the NAM might not help Tanzania address as it did in the early 1960s.
Of the 39 WAICO founding members through China, 10 are African countries, showing the selection they have made among the leading countries in AI. US still hasn’t included African countries as partner nations in the initiative.
The PwC report notes that AI will add 15 per cent to the world’s economy over the next decade, meaning countries like Tanzania, which have not advanced in AI technologies, will need to partner with AI-leading countries.
However, these partnerships may not yield better outcomes if governments and institutions are not prepared to act on their intent; a national policy to guide how AI can be adopted, developed, and regulated would be a better starting point.
This would encourage and pave the way for more experts in the field and for initiatives that advance industry innovation while leveraging ongoing academic research.
Gibson G. Mulokozi is a doctoral researcher at the Institute of Geography, University of Cologne, and a member of the Graduate School of Geosciences (GSGS) Steering Committee, with over six years of experience in social research. He is available at g.gmulokozi@uni-koeln.de. The opinions expressed here are the writer’s own and do not necessarily reflect those of The Chanzo. If you are interested in publishing in this space, please contact our editors at editor@thechanzo.com.
