The African Union (AU) is preparing to launch a continental credit rating agency in the second half of 2025 to address concerns that international credit rating firms unfairly disadvantage African nations.

The move, which was revealed in a report by the African Peer Review Mechanism (APRM) on February 11, is expected to challenge what many African leaders describe as arbitrary and biased assessments by dominant global rating agencies.

The APRM, which evaluates governance across AU member states and implements the New Partnership for Africa’s Development (NEPAD), has framed the creation of this new agency as a crucial step in reclaiming Africa’s financial autonomy. The report emphasized that the African rating agency would provide a more accurate and fair evaluation of sovereign credit risk, better reflecting the unique economic circumstances of African nations.

For decades, African governments and financial experts have criticized the methodologies employed by Moody’s, Fitch, and S&P Global Ratings, arguing that they consistently undervalue African economies. These agencies’ ratings determine the interest rates at which countries can borrow money and, in many cases, have led to excessively high borrowing costs for African nations.

The push for an African alternative gained significant traction in 2022 when Macky Sall, then-chair of the AU and president of Senegal, called for a new credit rating model to counter what he described as systemic financial injustices against African economies. Sall highlighted that in 2020, when economies across the world were reeling from the effects of the COVID-19 pandemic, African nations were disproportionately penalized.

“In 2020, when economies worldwide were struggling with the effects of COVID-19, 18 out of the 32 African countries rated by at least one of the major agencies saw their ratings downgraded. That’s 56% of African ratings being cut, compared to a global average of just 31% during the same period,” he said at the time.

It is believed that the assessment criteria used by these rating agencies incorporate subjective, non-economic factors, such as political stability, language, and governance models, which often result in negative outlooks for African nations. Research has suggested that up to 20% of the rating components assigned to African countries are based on factors unrelated to economic fundamentals. These assessments, critics say, artificially inflate borrowing costs, discourage foreign investment, and weaken African nations’ ability to finance key infrastructure and development projects.

The United Nations Development Program (UNDP) reinforced these concerns in an April 2023 report, which estimated that African countries had collectively lost out on $74 billion in potential financing due to unfair credit rating methodologies. The report found that the Big Three agencies rely on algorithms designed for traditional Western macroeconomic models, which often fail to capture the unique characteristics of African markets. The UNDP report also highlighted that credit analysts frequently base their assessments on prevailing investor sentiment rather than conducting deeper, localized analyses of economic conditions in African nations.

The upcoming African credit rating agency is being designed to function independently of political influence, with leadership drawn from the private sector. The AU has emphasized that credibility and transparency will be key to ensuring that the agency is widely accepted both within the continent and on the global stage.

According to the Africa Sovereign Credit Rating Outlook – 2024 Year-End Review, the agency will leverage experts based in Africa who have better access to real-time economic data and a deeper understanding of the unique financial landscapes across the continent. The report noted that “the Agency’s niche will primarily derive from its context-sensitivity, which will allow it to generate more comprehensive credit insights using competent experts based in Africa.”

Despite the enthusiasm surrounding this initiative, financial analysts caution that the success of the AU’s new credit rating agency will hinge on its ability to gain global recognition. Investors, financial institutions, and development banks still rely heavily on assessments from Moody’s, Fitch, and S&P when determining risk exposure. If these global institutions fail to recognize the African agency’s ratings as credible, its impact on borrowing costs may be limited.

Another challenge will be ensuring that the new agency remains free from political interference. To be taken seriously, it must establish strict independence from governments that might attempt to manipulate ratings for political gain. A perception of bias or lack of rigor could undermine the agency’s legitimacy in international markets.

Nevertheless, many African finance ministers and central bank governors believe that an African-led rating system could play a crucial role in correcting systemic financial imbalances. With greater control over its own credit assessments, it is believed that the continent could unlock new investment opportunities, reduce dependence on foreign financial institutions, and attract capital on more favorable terms.

Google’s AI leadership has dismissed concerns about DeepSeek, the Chinese startup that recently surged to prominence, with DeepMind CEO Demis Hassabis stating that Google’s artificial intelligence technology remains superior.

His comments, made during an all-hands meeting in Paris on Wednesday, reflect Alphabet’s confidence despite DeepSeek’s rapid ascent, which rattled U.S. markets after its AI chatbot topped Apple’s App Store, overtaking OpenAI’s ChatGPT.

At the meeting, Alphabet CEO Sundar Pichai read aloud a question summarizing employee concerns about DeepSeek’s success and the potential lessons for Google. DeepSeek gained attention after publishing a research paper claiming it had developed a powerful AI model at a fraction of the cost incurred by competitors. Employees were eager to understand the implications of this development on Google’s strategy as the company continued training future models.

Hassabis responded by suggesting that some of DeepSeek’s claims were “exaggerated.” He cast doubt on the startup’s reported cost efficiency, arguing that the total expenses of developing AI systems go beyond just the training phase. He implied that DeepSeek likely used significantly more hardware than it disclosed and may have leaned on existing Western AI technologies.

“We actually have more efficient, more performant models than DeepSeek,” Hassabis stated confidently. “So we’re very calm and confident in our strategy, and we have all the ingredients to maintain our leadership into this year.”

However, he acknowledged that DeepSeek’s achievements were notable, calling it “the best team I think I’ve seen come out of China.” He added that its progress carried “security” and “geopolitical” implications.

Several U.S. government agencies have already barred their employees from using DeepSeek, citing concerns over potential data privacy risks and the possibility of Beijing leveraging AI technologies for surveillance and cyber operations.

Google’s AI Policy Shift Sparks Internal Debate

Beyond DeepSeek, Google executives also faced employee concerns over a controversial revision to the company’s AI Principles, which quietly removed a long-standing pledge not to develop artificial intelligence for military weapons or surveillance applications.

Employees submitted a flurry of questions on the issue, prompting Pichai to address them directly. Reading an AI-generated summary of the concerns, he concluded with a pointed question: “Why did we remove this section?”

Kent Walker, Google’s president of global affairs, was tasked with explaining the shift. He acknowledged that he, along with Hassabis and senior vice president James Manyika, had led a “strategic reevaluation” of the company’s stance on AI ethics, beginning last year.

Walker traced the origins of Google’s AI Principles back to 2018 when the company declined to renew a controversial contract with the Pentagon known as Project Maven—a U.S. military initiative that used AI to analyze drone surveillance footage. At the time, backlash from Google employees led to the company’s firm opposition to AI-powered weaponry and intelligence gathering.

However, Walker suggested that Google’s strict prohibitions no longer align with today’s “more nuanced conversations” about artificial intelligence.

“An awful lot has changed in those seven years,” he said. “The technology has advanced to the point where it’s used in lots of very nuanced scenarios.”

Google’s move to soften its stance on AI in military and surveillance applications aligns with broader industry trends. Major tech companies, including Microsoft and Amazon, have actively pursued government contracts for AI-driven defense initiatives. Meanwhile, rising tensions between the U.S. and China have prompted calls for American firms to support national security efforts.

The decision, however, has reignited internal tensions within Google. The company has long grappled with employee activism, particularly when it comes to ethical concerns surrounding AI and government partnerships. The abrupt removal of the anti-weapons pledge has left many questioning whether Google is slowly reversing its past commitments to ethical AI development.