Many companies in Europe, the Middle East, and Africa (EMEA) significantly underestimate their dependencies on artificial intelligence (AI). According to a recent study conducted by IBM, only 10 percent of businesses in the region have a clear understanding of their AI-related interdependencies across vendors, models, and infrastructure. This figure rises slightly to 13 percent in Germany. The lack of clarity about these dependencies poses challenges in managing and securing AI systems within organizations.
The study highlights that switching primary AI providers or models proves difficult for many enterprises. A significant majority—73 percent in EMEA and 65 percent in Germany—find it challenging to transition away from their main AI vendor or model. Additionally, maintaining compliance with data residency and sovereignty requirements across regions is seen as difficult by 70 percent of respondents in both EMEA and Germany. These findings underscore the complexity involved in managing AI dependencies effectively.
Understanding the full scope of AI dependencies is crucial for maintaining control over critical operations. The concept of AI sovereignty does not imply complete independence but rather the ability to regain control when necessary. This involves identifying dependencies, managing them, and ensuring components can be replaced if needed. The AI stack encompasses data, models, infrastructure, and applications, making dependency management more intricate than traditional enterprise systems where dependencies often stop at infrastructure or application levels.
The risks associated with AI failures are substantial. According to the study, 81 percent of respondents in EMEA and 85 percent in Germany believe that a seven-day outage of their primary AI provider would lead to severe or critical consequences. On average, companies reported experiencing seven AI-related operational disruptions in the past two years, with causes ranging from provider services in EMEA to technical issues in Germany. These disruptions highlight the vulnerability of AI systems to both infrastructural problems and direct issues related to providers and models.
While many companies adopt a multi-vendor approach to AI, this strategy is often driven by organizational structures, regional mandates, and legacy complexities rather than a coherent strategy. Only 73 percent of respondents describe their AI environment as intentionally multi-vendor-oriented. However, achieving greater flexibility requires active management of AI environments. Without common standards for data, models, and security, complexity increases significantly. Interestingly, 71 percent of respondents in EMEA and 79 percent in Germany are willing to accept up to 20 percent higher costs for strategic flexibility.
The study advocates for selective sovereignty rather than a complete withdrawal from proprietary environments. This means exercising control selectively where it matters most for business operations. For instance, a transcription service might be assessed differently compared to a model influencing credit risk assessments, production decisions, or security processes. IBM suggests categorizing AI systems into three tiers: mission-critical systems requiring rapid data migration and tested fallback options, important but non-differentiating functions needing conscious management of dependencies, and commodity services that can be managed more flexibly.
Meanwhile, the debate around AI's role in journalism has intensified following incidents involving prominent figures. Mathias Döpfner, CEO of Springer, defended the use of AI-generated content, arguing against restrictions similar to those imposed on other technologies. He emphasized the importance of combining human and machine capabilities to maintain trustworthiness in journalism. His stance contrasts sharply with actions taken by other media outlets, such as the Frankfurter Allgemeine Zeitung (FAZ), which withdrew a guest contribution by Thuringia's Minister President Mario Voigt due to suspicions of AI involvement.
Voigt himself responded to allegations by asserting his personal responsibility for his speeches and writings, even if they were partially generated by AI. Critics argue that relying heavily on AI for political communication undermines authenticity and could provide fodder for right-wing critics. The controversy surrounding Voigt's potential use of AI has sparked broader discussions about the implications of AI in politics and media, raising questions about authorship, authenticity, and the future of human roles in public discourse.
The integration of AI into everyday tasks continues to evolve, with new concepts like "Agentic Commerce" exploring the feasibility of using autonomous AI agents for shopping. While promising, this approach raises concerns regarding product suitability, payment security, and data privacy. Current legal uncertainties further complicate the adoption of such technologies, highlighting the need for comprehensive regulatory frameworks to address emerging challenges.
As debates continue to unfold, the landscape of AI usage remains dynamic, reflecting both opportunities and challenges across industries and sectors. The ongoing dialogue between technology developers, policymakers, and users will shape how AI is integrated into society moving forward.
9 reports
Frankfurter Allgemeine (FAZ)Independent🔒CenterFactual 90Objective 7016 days agoMathias Döpfner, CEO of Springer, discussed his AI-generated article published in 'Welt' in response to the Frankfurter Allgemeine Zeitung (F.A.Z.) retracting a guest contribution by Thuringia's Minister President Mario Voigt, which was suspected of being written by AI. Döpfner argued that AI can be used both for research and writing, criticizing what he called the 'desperate attempt of the horse carriage lobby to ban the automobile.' He defended the use of AI in journalism and emphasized that it should not be judged solely on moral grounds but seen as an existential challenge. He also stated,
Bias read (Center): The article presents Döpfner's statements without overtly favoring one side of the debate over AI in journalism. It includes direct quotes from Döpfner and references the actions of the F.A.Z., providing a balanced view of the controversy.
Why these scores (Factual 90 · Objective 70): The article provides detailed quotes from Döpfner and contextualizes the debate around AI-generated content accurately. However, it uses emotionally charged language like 'verzweifelte Versuch der Postkutschen-Lobby' which introduces bias.
heise onlineIndependentCenterFactual 90Objective 6514 days agoA new study by IBM reveals that many companies in the EMEA region, including Germany, have a poor understanding of their dependencies on AI technologies. According to the survey, only 10% of companies in EMEA and 13% in Germany fully understand their interdependencies across providers, models, and infrastructure. The report highlights challenges in switching primary AI providers and meeting data residency requirements across regions.
Bias read (Center): The article presents findings from an IBM study without overtly favoring any political perspective. It focuses on corporate understanding of AI dependencies and does not include commentary or framing that suggests a political bias. The content remains factual and neutral in tone.
Why these scores (Factual 90 · Objective 65): The article thoroughly covers the debate over authorship and AI with philosophical references. However, it uses metaphorical and somewhat biased language that affects its neutrality.
HandelsblattIndependent🔒CenterFactual 85Objective 6519 days agoThe article title 'KI: In eigener Sache' translates to 'AI: On Its Own Terms,' suggesting the piece discusses artificial intelligence from a self-referential or autonomous perspective. However, no further content is provided in the text, making it impossible to determine the specific focus or details of the article.
Bias read (Center): No explicit political stance or biased language is present due to the lack of detailed content. The available information does not indicate a clear ideological leaning.
Why these scores (Factual 85 · Objective 65): The article mentions the topic but lacks specific details or quotes. The title suggests self-interest but does not provide enough information to assess factual accuracy or objectivity effectively.
Die WeltIndependent🔒CenterFactual 85Objective 6021 days agoThe article discusses the debate surrounding Mario Voigt in the context of artificial intelligence, referencing 'Die Maschinen-Stürmer' from the Frankfurter Allgemeine Zeitung (FAZ), suggesting a futuristic perspective on AI developments.
Bias read (Center): The article appears to present a neutral discussion on an ongoing debate without overtly favoring any particular side or using biased language.
Why these scores (Factual 85 · Objective 60): The article presents the controversy around Mario Voigt's article accurately but uses hyperbolic language such as 'Willkommen im Jahr 2026!' which skews the tone towards sensationalism.
heise onlineIndependentCenterFactual 80Objective 7521 days ago AI is going shopping.The article discusses 'agentic commerce,' which involves using autonomous AI agents for shopping tasks. It raises questions about the risks and challenges associated with this technology, such as ensuring product suitability, payment security, potential misunderstandings leading to impulsive purchases, and data privacy concerns like the 'glass customer' issue. The piece explores perspectives on the concept and highlights current legal uncertainties surrounding it.
Bias read (Center): The article presents a balanced discussion of agentic commerce without overtly favoring any particular perspective. It outlines both the potential benefits and risks of the technology while acknowledging existing legal uncertainties. There is no detectable ideological framing or biased language.
Why these scores (Factual 80 · Objective 75): The article discusses agentic commerce with clear explanations of risks and uncertainties. It remains largely objective while acknowledging the complexities involved in AI shopping.
Frankfurter Allgemeine (FAZ)Independent🔒CenterFactual 80Objective 6017 days ago The authorship debate: If AI Voigt doesn't cost the jobThe article discusses the debate around authorship in the context of artificial intelligence, referencing literary theory and philosophical concepts such as those of Roland Barthes and Michel Foucault. It critiques the current discourse on AI's impact on human identity and creativity, using the example of Mario Voigt, a German writer, whose potential loss of authorship due to AI-generated content is highlighted. The piece reflects on the broader implications of AI on human autonomy and cultural expression.
Bias read (Center): The article engages with philosophical and cultural themes rather than taking a clear ideological stance. It references theoretical perspectives without overtly favoring one side of the AI debate. The tone is critical but balanced, avoiding direct political commentary or biased language.
Why these scores (Factual 80 · Objective 60): The article touches on AI and migration topics but lacks depth. Its title implies a humorous angle, yet it doesn't provide sufficient factual detail or maintain complete neutrality.
SternIndependentCenterFactual 75Objective 6019 days agoThe article discusses the concept of 'remigration' and explores the idea of using artificial intelligence to determine where individuals should be relocated, potentially even sending them to places like Kansas.
Bias read (Center): The article appears to be discussing a hypothetical or conceptual topic related to AI-driven migration decisions without taking a clear stance or showing bias toward any particular political ideology. The framing remains neutral, focusing on the technological aspect rather than making value judgents
Why these scores (Factual 75 · Objective 60): The article briefly mentions AI in media but focuses more on other topics. It maintains some level of factual accuracy but lacks balance in its approach to the subject matter.
taz – die tageszeitungIndependentCenter19 days agoThe article discusses Germany's progress (or lack thereof) regarding rights for LGBTQ+ individuals, the potential impact of a World Cup group stage exit as a statement, and the use of AI in media. It includes an interview with Friedrich Küppersbusch discussing various topics including football tournaments, Trump's age, and the use of AI by media outlets such as the Tagesspiegel and FAZ. The article also touches on the controversy surrounding AI-generated content and its implications for journalism.
Bias read (Center): The article presents a discussion on multiple topics without overtly favoring any particular political stance. While it mentions controversial issues like AI in media and LGBTQ+ rights, the framing remains balanced and does not exhibit clear ideological bias.
taz – die tageszeitungIndependentCenter23 days ago AI speeches by Mario Voigt: pathetic thoughtlessnessThe article discusses allegations that Mario Voigt, the Minister-President of Thuringia and member of the CDU, has been using AI to generate his speeches. It references an analysis by the text recognition program Pangram, which claims that over 50% of Voigt's speeches were generated by AI, with three appearing fully AI-generated. One such speech was reportedly delivered on Holocaust Remembrance Day in January 2025.
Bias read (Center): The article presents factual information based on an external analysis tool (Pangram) without overtly biased language or selective sourcing. The content focuses on the technical claim regarding AI-generated speeches rather than taking a clear ideological stance.
★
Keep the news honest.
ObjectiveNews is reader-funded and ad-free — we show you the bias instead of hiding it. Support independent journalism for €5/month.
Become a Supporter
Germany
