Bertelsmann Stiftung

The Bertelsmann Stiftung is a German private operating foundation established on 8 February 1977 by Reinhard Mohn — the Bielefeld-born media-industry executive who built the family-owned Bertelsmann publishing company into a global media group — and headquartered in Gütersloh, North Rhine-Westphalia at Carl-Bertelsmann-Straße 256, 33311 Gütersloh, in continuous use since 1991. The foundation is a non-profit foundation under North Rhine-Westphalian private law, designed by Mohn as an operating foundation — one that develops and oversees its own projects rather than primarily disbursing grants to external organisations — and structurally resourced through dividends on a substantial capital-share stake in Bertelsmann SE & Co. KGaA, the family media group whose foundations together hold 80.9 per cent of the company's capital shares (Bertelsmann Stiftung, Reinhard Mohn Stiftung, BVG-Familienstiftung, and BVG-Stiftung), without voting rights — the Mohn family retains the remaining 19.1 per cent and exercises voting control through Johannes Mohn GmbH. The current executive board comprises Dr. Brigitte Mohn as chair, Prof. Dr. Daniela Schwarzer, and Wilhelm-Friedrich Uhr, with the supervisory Board of Trustees chaired by Bodo Uebber.

Origin and founding

Reinhard Mohn established the foundation in 1977 with a 100,000 Deutsche Mark initial endowment, formally approved by the State Ministry of North Rhine-Westphalia on 14 March 1977, and recruited Hans-Dieter Weger in 1979 as the first managing director to operationalise the operating-foundation model the founder had designed. The defining structural move came in 1993, when Mohn transferred the majority of the family's Bertelsmann Group capital shares to the foundation — separating capital ownership from voting control — with the dual purpose of ensuring corporate continuity at Bertelsmann SE and routing the family's philanthropy through the foundation rather than through direct grantmaking. The transfer also produced a tax saving on the order of €2 billion in inheritance and gift taxes, which has remained a recurring point of public-policy criticism of the foundation's non-profit status — most prominently in independent legal-expert opinions commissioned in 2009 concluding the foundation "no longer met non-profit requirements" and in Thomas Schuler's 2010 book Bertelsmannrepublik Deutschland: Eine Stiftung macht Politik — though the foundation has retained its non-profit recognition under German foundation law throughout. The foundation's quoted founding self-framing, attributed to Mohn, is that "The Bertelsmann Stiftung is a place where we look across political boundaries to the future, so we can develop new impulses for change".

Programmatic structure and resourcing

The foundation operates six named programme areas — Education, Democracy, Society, Health, Culture, and Economy — with approximately 330 employees across its Gütersloh headquarters and additional offices in Berlin (Kommandantenhaus and Werderscher Markt), Brussels (European Quarter), Barcelona (Fundación Bertelsmann, mid-1990s), and Washington, DC (Bertelsmann Foundation North America since 2008). Financial scale across the foundation in 2023 was €163 million in revenue, €75.6 million in total expenditures, and €38.7 million in direct programme spending, with cumulative grantmaking since 1977 reported at €1.9 billion — placing Bertelsmann Stiftung among the largest private foundations in Germany by both operating scale and cumulative giving. The operating-foundation model means that much of the spend is on in-house project work — recurring initiatives include the Neue Stimmen international opera competition (1987–present, launched by Liz Mohn), the Centre for Higher Education (CHE, 1994, co-founded with the German Rectors' Conference and held 90 per cent by the foundation), the Reinhard Mohn Prize (formerly the Carl Bertelsmann Prize, 1988–present), and the Weiße Liste patient-information internet portal — rather than the open-call external grantmaking model of US private foundations like Ford or MacArthur.

AI-good footprint

The foundation's AI-good footprint is built around a single sustained project line and its associated outputs.

The first thread is the Ethics of Algorithms project — launched in 2017 under Carla Hustedt's leadership — and renamed reframe[Tech] – Algorithms for the Common Good on 6 September 2022 under an expanded team. The post-2022 team, currently led by Dr. Sarah Fischer as Senior Expert with Dr. Felix Sieker and Asena Soydaş as Project Managers, structures its work around two programmatic priorities — Mitigating Algorithmic Risks and Harnessing Potential for the Common Good — and a four-mission strategy update covering the amplification of underrepresented voices in algorithmic-policy discourse, evidence-based counterweights to excessive AI optimism, sectorally specific progress over speculative hype scenarios, and the prevention of digital-transformation-driven inequality. The project's New New Fellowship has backed 12 projects from 9 European countries on inclusive digital futures, and the dedicated blog at ethicsofalgorithms.org (preserved through the rename) remains the project's public-facing publication channel.

The second thread, anchoring the project line in the corpus's algorithmic-accountability publication slate, is the Automating Society mapping series — co-published by AlgorithmWatch and the Bertelsmann Stiftung. The 2020 edition, published on 28 October 2020, is the corpus's anchor publication for the series: a 297-page joint mapping of ADM systems across sixteen European countries plus the EU institutional level, edited by AlgorithmWatch's Fabio Chiusi under Matthias Spielkamp's editorial direction, and placed on the Bertelsmann side inside the foundation's then–Ethics of Algorithms programme. The 2019 12-country edition preceded it; a Swiss Edition followed on 28 January 2021, extending the methodology to Switzerland. The Bertelsmann–AlgorithmWatch relationship is best read as a sustained project partnership rather than a budget-funding relationship — AlgorithmWatch's own transparency reporting names six >10%-of-budget funders (Alfred Landecker Foundation, BMBF, BMUV, Deutsche Postcode Lotterie, Schöpflin Stiftung, Stiftung Mercator) and does not list Bertelsmann among them — but the joint-publication line is the single most visible joint output between any German foundation and AlgorithmWatch on the corpus's record.

The third thread is the Algo.Rules nine-principle framework for the ethical development and use of algorithmic systems, published in April 2019 by the Bertelsmann Stiftung and the Berlin think tank iRights.Lab through a multi-year participatory process bringing together IT experts, humanities and law representatives, and individuals from politics, civil society, and business. The principles include explicit human-rights anchoring (the rules are positioned to strengthen, not undermine, individual and collective freedoms and rights), enforceability requirements for the regulations protecting those norms, and concrete operational guidance for IT and software companies including success-factor analysis and practical checklists for developers and executives. Carla Hustedt led the development of Algo.Rules during her tenure at the Ethics of Algorithms project before her April 2021 move to direct Stiftung Mercator's Centre for Digital Society — a personnel-and-leadership transfer that is the corpus's clearest example of staff continuity between the corpus's German foundation entries on the AI-good file, with Algo.Rules and the post-rebrand European AI & Society Fund Steering Committee co-chair seat both held by the same individual across two consecutive employers.

A fourth ancillary thread is the foundation's eupinions survey work on algorithmic attitudes — What Europe Knows and Thinks About Algorithms (February 2019), authored by Viktoria Grzymek and Michael Puntschuh, is the most-cited example, an early quantitative cross-European population survey reporting that less than half of Europeans were familiar with algorithms, eight per cent had substantive knowledge of how they worked, nineteen per cent associated algorithms with "manipulation" against twenty-seven per cent with "efficiency", and seventy-four per cent wanted stricter controls on their use. The survey has been cited by subsequent European AI-policy research as one of the earliest quantitative baselines for algorithmic public-opinion data on the continent.

Position in the corpus funder slice

Within the funder slice of this corpus Bertelsmann Stiftung is the third German private foundation entry, joining Stiftung Mercator and Robert Bosch Stiftung, and the corpus's first former European AI & Society Fund partner — distinct from each of the existing nineteen entries on three structural axes. It is the first operating foundation in the funder slice — designed by Reinhard Mohn explicitly as a project-running foundation rather than a primarily-grantmaking foundation, a structural type distinct from the conventional endowed-grantmaker model that all other in-corpus funders follow. It is the second operating-company-shareholder-anchored corporate foundation in the corpus after Robert Bosch Stiftung — both are financed through dividends on substantial ownership stakes in the founding family's operating company rather than through a separate investment endowment, but the Bertelsmann Stiftung holds its 80.9 per cent stake jointly with three related Mohn-family foundations without voting rights (the 1993 transfer's defining feature), where Bosch holds a 94 per cent stake directly in Robert Bosch GmbH. And it is the corpus's first former EAISF partner — a structurally distinct slot from Mercator's founding-partner-and-Steering-Committee-co-chair role and Bosch's later-joining contributor-tier role: Bertelsmann Stiftung appears on the EAISF current-partners page under "Our previous partners" alongside Isocrates Foundation, the two former contributors who are no longer funding the vehicle out of EAISF's "20 foundations have contributed since 2020" cumulative-partner framing.

The relationship to Stiftung Mercator is the closest sibling pairing in the corpus. Both are German national foundations with Continental European AI-good footprints anchored in a flagship algorithms-and-society project — Bertelsmann's Ethics of Algorithms/reframe[Tech] line and Mercator's Digital Society programme — and the two foundations are linked through Carla Hustedt's personal trajectory from leading the former until April 2021 to directing the latter's Centre for Digital Society from that point. The functional split, however, is sharp: Mercator is the current EAISF founding partner, the Steering Committee co-chair, and the multi-year >10%-of-budget direct funder of AlgorithmWatch's operating capacity; Bertelsmann is the former EAISF partner whose own AI-good investment has run through in-house project-running (reframe[Tech]), participatory norm-setting (Algo.Rules), survey production (the eupinions algorithms work), and project partnership (the Automating Society series with AlgorithmWatch) rather than through direct grant-line funding of in-corpus organisations.

The foundation's relevance to the corpus's grassroots-democratic AI-good mission runs primarily through three layers: the Automating Society publication line, which supplied the recurring empirical evidence base on which European civil-society AI Act advocacy was built (and which the corpus tracks under Automating Society Report 2020); the Algo.Rules normative framework, which influenced the European algorithmic-accountability discourse during the run-up to the EU AI Act and which has been cited in subsequent practitioner-facing AI-ethics work; and the Carla Hustedt staff-trajectory thread that links Bertelsmann's earlier AI-good project leadership to Mercator's continuing EAISF Steering Committee co-chair role. The foundation's broader public-policy record — its programme spend is concentrated in education, demographic change, healthcare, and European integration rather than in the specific grassroots-democratic AI-good vertical — and its long-running history of structural criticism on neoliberal-influence and conflict-of-interest grounds (Schuler 2010; the LobbyControl and taz critiques recorded in the foundation's public-controversy literature) place it as a structurally different funder type from the human-rights-grantmaking and pooled-philanthropy vehicles that dominate the rest of the corpus's funder slice. The funded_orgs field is left empty pending direct grant-to-corpus-entity records, matching the convention used elsewhere in the funder slice for indirect-resourcing and project-partnership relationships.