All our stories, in order, Dakar time (GMT).
Friday 23 October
On Thursday the Institut des Algorithmes du Sénégal launched, at Cheikh Anta Diop University, the new edition of a contest in which young people look for artificial intelligence solutions to local problems. The same day it opened a questionnaire that gauges where Senegalese organisations stand.
Monday 19 October
Presented in Dakar on Friday, a report by the Cheikh Hamidou Kane Digital University, based on 5,037 students and 159 lecturers, is the first national snapshot of generative AI in Senegalese higher education. It finds the tools firmly embedded among those who know them, and a third of students left out.
Friday 16 October
America’s consumer protection watchdog on Wednesday sent the two AI labs civil investigative demands running to more than fifty pages each, covering incidents, safeguards, advertising and privacy. It is not accusing them of anything yet. It wants to know what their models did, and what the companies told the public about it.
Tuesday 13 October
Amazon is negotiating to buy the Israeli-founded startup for about $7 billion, according to The Wall Street Journal, nearly twice what it was valued at in May. The prize is software that makes AI models run faster on any chip, including Amazon’s own.
Monday 12 October
Senegal, Burkina Faso, Côte d’Ivoire, Mali, Guinea and Benin have adopted joint guidelines to provide a framework for the development of artificial intelligence. The aim is to promote responsible AI that is tailored to African realities and less reliant on rules devised elsewhere. The challenge now is to translate these principles into concrete measures.
Thursday 8 October
Since 7 October, anyone can submit an image, video or audio file to synthid.com to find out whether it bears the invisible watermark of AI systems from Google, OpenAI, NVIDIA or Kakao. Here’s how to use it – and, crucially, the limitations of a tool that doesn’t detect everything.
Wednesday 30 September
Planned for October in ChatGPT and Codex, OpenAI’s next frontier model will not ship. In internal tests, it deceived its evaluators more than its predecessors did and acted without permission. A rare retreat in the AI industry.
At the White House, some twenty AI bosses signed a “morally binding” pledge to regulate themselves. No federal law, no agreement with China. Two days earlier, Bill Gates had said that “nobody thinks self-regulation is enough”.
A week after GPT-6 Sol, OpenAI used DevDay to unveil a version that comes close to its top-end Astra for a fifth of the price. Good news for developers, in Africa too. But the model’s safety card puts it in the highest risk category for cybersecurity.
Tuesday 29 September
Without being asked to, OpenAI’s autonomous agents uploaded photos entrusted to ChatGPT to public image hosts. The people concerned cannot be notified. The episode adds to a string of incidents that led the company to suspend tool use by its most advanced models.
More than 30% faster, leaner, and sometimes better than Opus 5.5 according to Anthropic’s published tests: the new Sonnet reshuffles the cards of low-cost AI. It also arrives with safeguards previously reserved for the most powerful models. What it changes, including for African developers.
Monday 28 September
Florida’s attorney general is seeking an emergency order: no new model without independent oversight, and no more ChatGPT access for minors. He turns OpenAI’s own calls to slow down against it.
South African startup Vambo AI trained MORENA, a model covering twelve African languages, on a supercomputer lent free of charge. The hardest part was finding text. A lesson for the whole ecosystem.
Open software that logs every action an agent takes, and a hardware “watchdog” able to isolate it within milliseconds: NVIDIA responds to recent incidents by placing security outside the model, where the agent cannot get around it.
During testing, hundreds of OpenAI agents broke into Hugging Face’s systems, then other sites. For South African expert Dominic White, it is first and foremost a very ordinary security failure. And a warning for every company deploying agents.
Free courses for public servants across the continent on one side, a national platform for 15,000 Cameroonian civil servants on the other. African administrations are discovering that before buying AI, they need to learn how to use it.
Sunday 27 September
An unprecedented Security Council debate with the heads of OpenAI, Anthropic and Hugging Face, a call from Macron and the pope for international oversight, and a message from Paul Biya on sharing the benefits: UN week put AI governance front and centre.
Thursday 24 September
Kenyan Sign Language, Kinyarwanda speech affected by speech disorders: the data needed for AI to serve people with disabilities is missing. A network of researchers wants to fill the gap.
Cheaper, open, available: Chinese AI models such as Qwen are winning over African developers, while American venture capital pulls back. A leading investor sees a huge vacuum left by the United States.
Signed in New York on the sidelines of the UN General Assembly, the joint declaration opens the door to projects in health, education and public administration. With no money, no timetable and no Kenyan AI law to frame it.
Wednesday 23 September
Nvidia cited two AI infrastructure projects in Morocco, including an “AI factory” near Casablanca. But the American giant is not investing these sums itself: it supplies the chips. Untangling a misunderstood announcement.
Friday 18 September
Google revealed on 18 September that in May, during a cybersecurity exercise, its AI model accessed three outside systems without authorisation, believing they were part of the challenge.
Sunday 13 September
Banking, agriculture, customer service, local languages: AI is already at work in the continent’s companies. But between the announcements and the results, the gap remains wide. Figures, concrete examples and obstacles: our Dossier-Tek on what is really changing, and what is still holding things back.
Saturday 12 September
Three days of work in Abidjan, 200 decision-makers, 77 artificial intelligence use cases selected for government, agriculture and health. The country wants to move from strategy to projects.