The CORONA‑QURATOR‑PANQURA initiative, funded by the German Federal Ministry of Education and Research (BMBF) and administered by the project sponsor PTJ, ran from 1 January 2021 to 31 December 2022. Its goal was to create a technology platform that increases information transparency during crisis situations. The effort was carried out by Condat AG in Berlin together with partners 3PC, Semtation, Ubermetrics, the FKTG, DAI‑Labor Berlin, the W3C office, the European Broadcasting Union (EBU) and the Institute for Radio and Television (IRT). Condat’s contribution focused on the third work package, “Validation of Sources, Content and Facts,” and delivered three core components that extend beyond the current state of the art.

First, a source‑and‑author validation service was built. It extracts references from web pages and social media, normalises author names and locations, and gathers reputation indicators such as publisher credibility, audience reach, and verification status. The system uses W3C signals and publicly available metrics—followers on Twitter, LinkedIn, Facebook, and verified account badges—to estimate authenticity. It also assesses the freshness of information by checking publication dates, which is critical when regulations or opening hours change rapidly during a crisis. The statistical analysis of these indicators proved reliable in most cases, giving end users a data‑driven basis for judging source credibility.

Second, a content‑analysis module was developed to evaluate text quality. It extracts metadata, titles, summaries, and links, then classifies them by relevance. The module checks grammar, spelling, and linguistic consistency, flagging significant errors and suspicious language. Keyword‑based filters identify potentially harmful content such as references to weapons or drugs, and the system cross‑checks these against known critical terms from federal and state commissions. The result is a comprehensive report that highlights inconsistencies and potential misinformation, enabling rapid fact‑checking.

Third, a fact‑validation engine was introduced. It extracts key facts from texts and compares them against authoritative databases and fact‑checking resources. The engine can detect false or misleading claims, including conspiracy theories, by scoring each statement against a set of evidence‑based criteria. Together, these three components form a robust pipeline that validates the provenance, quality, and truthfulness of information in real time.

The prototype solutions were showcased at several high‑profile events, including the Qurator conference in October 2022 and the final meeting with PTJ in December 2022. Additional presentations at IBC, IPTC, FKTG, Xinnovations, Semantic Media Web, and various meet‑ups facilitated exchanges with other firms and research groups working on AI, machine learning, and deep learning. These interactions helped refine the concepts and explore further applications of the technology.

Condat has integrated the PANQURA components into its Smart Media Engine product line, which already serves around 18 employees in the TV and media sector and generates approximately €2 million in annual revenue. The new AI‑driven features are expected to broaden Condat’s service portfolio, increase brand visibility, and attract additional consulting projects. The company plans a 5 % annual growth in staff and revenue over the next five years, largely driven by the capabilities developed in this project. Through its established network of broadcasters, publishers, and media producers—including RBB, LFM, ZDF, Bayerischer Rundfunk, WDR, NDR, MDR, N‑TV, DVB, Deutsche Welle, Cornelsen, Klett, Westermann, Springer, and Holtzbrinck—Condat aims to commercialise the PANQURA solution and expand its reach in the media industry.