algorithmwatch

algorithmwatch, 15 days ago to random

PR-Managerin für Berlin gesucht*

Für unseren Berliner Standort suchen wir nach einer/einem PR-Manager*in mit einem Herz für Menschenrechte, Gerechtigkeit, Demokratie und Nachhaltigkeit beim Einsatz von KI und Automatisierung - Bewirb Dich bis 22. Mai, wir freuen uns, Dich kennenzulernen!

https://algorithmwatch.org/de/job-managerin-communications-pr/

algorithmwatch, 2 months ago to ai

💰 What is the money for? The company that manages Austria’s health insurance system’s technical infrastructure passed a massive tender for #AI services. They want to give away €52m, but the plan is shrouded in secrecy, even though the company’s track record urges caution. AlgorithmWatch tried to get more details about the intended use areas, but requests were left unanswered. Our fellow @fin tells the whole story! ⤵️
https://algorithmwatch.org/en/austrias-social-security-invests-massively-in-ai/

algorithmwatch, 3 months ago to random

💡 Did you know? More than half of the companies in Germany use algorithm hiring tools, to source, search, and select candidates! Leading to job applicants with the best skills and experiences not necessarily getting the job interview.

This was the result of an experiment held in Germany and the UK, by a team of researchers around @paksy for the Horizon EU FINDHR project, where we are working on preventing discrimination through AI in job selection! ⤵️
https://findhr.eu/wp-content/uploads/2024/01/FINDHR-Expert-Report_by-Paksy-Plackis-Cheng-et-al.pdf

algorithmwatch, 3 months ago to random

We want to hear from you! In the Horizon EU project @findhr we are working on tools that prevent discrimination through AI in job selection.Together with our partners we have been focusing on three research topics: an algorithmic impact assessment, an equality monitoring protocol & a software development guide. Our question for you: What do you think about our proposals? Is something missing? We want to learn from our community, please leave a comment on the website. ⬇️
https://findhr-feedback.upf.edu/processes/feedback-request-01?locale=en

algorithmwatch, 3 months ago to ai

Fighting against #AI: Automated systems can contribute to child sexual abuse, denying people their social benefits, or blocking organizations' online presence. Our fellows @mathildesaliou and Alina Yanchur talked to Europeans fighting back against AI systems gone astray. Their stories show that we have a long road ahead of us before effective remedy mechanisms are in place. Read the whole story! ⤵️
https://algorithmwatch.org/en/europeans-push-back-against-rogue-ai/

algorithmwatch, 3 months ago to ArtificialIntelligence

🤔 How are people discriminated by #algorithms? In our Horizon Europe project @findhr we are working on uncovering discrimination through #AI in job selection. For this, the author Jie Liang Lin has taken on the issue of algorithmic hiring and systemic discrimination against migrant tech professionals in her latest research. We have gathered some quotes from people, challenged with curating their digital profiles to make their competences visible. Find out more! ⤵️
https://findhr.eu/wp-content/uploads/2024/02/FINDHR-Expert-Report_by-Jie-Liang-Lin.pdf

image/png
image/png
image/png

algorithmwatch, 3 months ago to random

📣 We are looking for you! If you are a passionate artist or designer, who is interested in visualizing our project and the topic of algorithmic discrimination, you have come to the right place. In our Horizon Europe project @findhr we are working with our partners on uncovering discrimination through algorithms in the recruiting process. Don’t miss out on the chance to work in human rights & promote the work through our large network all over Europe. Find out more! ⤵️

https://findhr.eu/open-call-for-artists-visualization-on-discrimination-in-algorithmic-hiring/

algorithmwatch, 3 months ago to Bulgaria German

Morgen stimmen die #EU-Mitgliedsstaaten über den #AIAct ab – das Gesetz, mit der die EU eine rechtliche Grundlage für die Entwicklung und den Einsatz von Künstlicher Intelligenz schaffen will.

Dabei steht viel auf dem Spiel, wenn es um die Frage geht, wie #KI klimafreundlicher werden kann.

Mehr dazu in unserem Explainer zu KI und #Nachhaltigkeit. ⬇️

https://algorithmwatch.org/de/nachhaltigkeit-ki-erklaert/

algorithmwatch, 4 months ago to random

🎧 Irreführende Antworten, erfundene Umfragewerte und ausgedachte Skandale: Der WDR 5 berichtet in einem Interview mit unserem Geschäftsführer Matthias @spielkamp über unsere Studie zu den #Wahlen in Hessen, Bayern und in der Schweiz. Bei dieser haben wir die Antworten des Microsoft #KI-Chatbots "Copilot" systematisch untersucht. Jetzt das Interview anhören und herausfinden, warum KI-Chatbots nicht die besten Informationsquellen darstellen. ️⤵️
https://www1.wdr.de/mediathek/audio/wdr5/wdr5-toene-texte-bilder-interviews/audio-ki-chatbots-wahrhaftigkeit-in-gefahr-100.html

algorithmwatch, 4 months ago to random

📣 We have great news! For our algorithmic accountability reporting fellowship, seven journalists and researchers have joined us to work on uncovering discrimination arising from the use of automated decision-making systems in Europe's financial sector. Welcome @pjarandia, Maryam Ahmed, Yasir Gökce, Sara Kezia, Maja-Lee Voigt and Mathana. Follow us and stay tuned for their exciting work! ⤵️
https://algorithmwatch.org/en/fellows-investigate-discrimination-in-financial-sector/

algorithmwatch, 4 months ago to random

Die neuste Schlagzeile von @tonline wundert uns nicht! Wie die Suchmaschine #Bing spuckt auch der darauf basierende #Chatbot Microsoft Copilot Falschinformationen aus. Das konnten wir auch in unserer Studie zu den Wahlen in Hessen, Bayern & in der Schweiz zeigen, die aufdeckt, wie unzuverlässig Informationen von KI-Chatbots sein können. ⬇️
https://algorithmwatch.org/de/schlussbericht-microsoft-bing-chat/

algorithmwatch, 4 months ago to random

🎵 Pre-save links are everywhere: Artists use them to promote their unreleased tracks. But this comes with a cost: They’re forced to become #influencers and marketing experts. Read our fellow Mathilde Saliou’s article and learn why music is a #data business. ⬇️
https://algorithmwatch.org/en/pre-saves-algorithmic-marketing-music/

algorithmwatch, 5 months ago to random German

Kreativität und Leidenschaft für Kampagnenplanung sind deine Stärken?

Dann bist du bei uns genau richtig! Setzt dich gemeinsam mit uns für Menschenrechte ein und werde Teil unseres tollen Teams in Berlin.

Als Senior Campaign Manager*in bist du für die Entwicklung und Leitung unserer Kampagnen verantwortlich. Bewirb dich jetzt!

👉🏻 https://algorithmwatch.org/de/job-senior-campaign-managerin/

algorithmwatch, 5 months ago to ai

Let’s create a world where technology & #AI is used to protect and benefit all human beings. ✊✊🏼✊🏿

Support us so we can continue fighting for human rights by protecting our rights for privacy and standing up against any sort of discrimination through algorithms and hashtag#ArtificialIntelligence!

#InternationalHumanSolidarityDay 🧡

👉🏻 Help us and become part of our community of supporters: https://algorithmwatch.org/en/supporting-member/

algorithmwatch, 5 months ago to random

📣 On today's #InternationalMigrationDay, let’s shed light on EURODAC – the EU database for the registration of asylum-seekers. EU member states want to turn it into a surveillance tool that treats people seeking protection like criminal suspects!

Deportations, detention & racialized criminalization: This will increase immensely due to the reform of the EU database! We need to protect human rights & stop the expansion of #EURODAC!

Share our joint statement now. ⬇️
https://edri.org/our-work/civil-society-calls-for-an-end-to-the-expansion-of-eus-eurodac-database/

algorithmwatch, 5 months ago to random

🆕 Out now: New report on Microsoft Bing Chat.

Bing Chat, the AI-driven chatbot on Microsoft’s search engine Bing, makes up false scandals about real politicians and invents polling numbers. Microsoft seems unable or unwilling to fix the problem.

These findings are based on a joint investigation by AlgorithmWatch and @aiforensics_org , the final report of which has been published today.

Read our full report: https://algorithmwatch.org/en/study-microsofts-bing-chat/

algorithmwatch, 5 months ago to random

After a negotiation marathon on its Artificial Intelligence Act, EU Member States put the screws on the EU Parliament to put national security and industry interests before the protection of people’s rights.

Having been negotiating for over 20 hours, sleep-deprived EU lawmakers were pressuring each other today to close an unacceptable deal – on some of the most fundamental implications of AI on people and society.

https://algorithmwatch.org/en/ai-act-drama/

algorithmwatch, 5 months ago to random German

Der Europäische Gerichtshof hat entschieden: Das #Schufa-Scoring muss als automatisierte Entscheidung eingestuft werden.

Daraus leitet sich unter anderem eine Stärkung der Aufklärungs- und Informationspflichten der Schufa gegenüber den Verbraucher*innen ab, wie @spielkamp von AlgorithmWatch im Interview mit @ntv erklärt.

📺 Zum Beitrag: https://www.n-tv.de/mediathek/sendungen/teleboerse/Teleboerse-von-10-45-Uhr-article9834416.html

algorithmwatch, 5 months ago to random German

"Trotz der drohenden Verwässerungen wünscht sich Kilian Vieth-Ditlmann von AlgorithmWatch nicht, dass die Trilog-Einigung am Mittwoch scheitert.

'Ein erfolgreicher Abschluss hängt aktuell am seidenen Faden. Und kommt es nicht zu einem tragfähigen Kompromiss, ist die Verabschiedung der gesamten #KI-Verordnung in Gefahr', so Vieth-Ditlmann. 'Es wäre aber verantwortungslos, das Gesetzgebungsverfahren jetzt noch scheitern zu lassen.'“

📰 @netzpolitik_feed #AIAct : https://netzpolitik.org/2023/ki-verordnung-trilog-einigung-haengt-am-seidenen-faden/

algorithmwatch, 6 months ago to random

Have you heard of #PlantNet or #PictureThis? 🪴 They are plant-identifying apps, that help amateur botanists with their hobby. However, they have serious limitations especially for students in this field. Read more in our fellow Pablo Jiménez Arandia latest story! ⤵️
https://algorithmwatch.org/en/plant-identifying-apps/

algorithmwatch, 6 months ago to Bulgaria

Apparently, adoption of the #AIAct as a whole is at risk because the #EU Council and Parliament are unable to reach a common position on generative AI. This is highly irresponsible, as it threatens effective prevention of harms caused by AI-driven systems in general.

An op-ed by @Nhelberger , Bettina Berendt, Nick Diakopolous, angelamueller@mstdn.social , @LaurensNaudts , spielkamp@mastodon.social , Sophie Weerts & Kilian Vieth-Ditlmann.

https://algorithmwatch.org/en/op-ed-generative-ai-ai-act-trilogue/

algorithmwatch, 6 months ago to random

KI-basierter #Gesichtserkennung gelingt es praktisch jedes Gesicht zu identifizieren. 👀

Welche Gefahren sind damit verbunden?
Das erklärt unser stellvertretender Leiter des Policy- & Advocacy-Teams, Kilian Vieth-Ditlmann, in der @deutschlandfunk Reihe "#KI verstehen".

Jetzt reinhören! 🎧
https://www.deutschlandfunk.de/gesichtserkennung-macht-ki-uns-zu-glaesernen-buergern-dlf-f5b06014-100.html

algorithmwatch, 6 months ago (edited 6 months ago) to random

Automated image generators are often accused of spreading harmful stereotypes. We wanted to find out how biased different tools are and compared the results when using the same prompt across MidJourney, Kandinsky and Adobe Firefly.

Take a look at the results: https://algorithmwatch.org/en/image-generators-stereotypes-diversity/

algorithmwatch, 7 months ago to random

Langsam wird es knapp für die Versprechen im #Koalitionsvertrag! ⏰ Die #Ampel muss jetzt handeln. Daher fordern wir gemeinsam mit @Freiheitsrechte, Wikimedia Deutschland, @reporterohnegrenzen_feed & Open Knowledge Foundation Deutschland in unserem Bündnis F5:

➡️ Zivilgesellschaft einbinden
➡️ Grundrechtskonforme Sicherheitsgesetze
➡️ Bundestransparenzgesetz und Recht auf Open Data
➡️ Wirksame Kontrolle von Plattformen & Algorithmen

Mehr Infos im @netzpolitik_feed Artikel.
https://netzpolitik.org/2023/halbzeitkritik-diese-vier-vorhaben-muss-die-ampel-endlich-anpacken/