AI for Animals #8: Retrospective on AIA Bay Area 2025
April 15, 2025
Welcome to the eighth edition of AI for Animals, the newsletter bringing you the latest news on AI’s potential impacts on animals.
First, an announcement:
AI, Animals, & Digital Minds 2025 will bring together leading voices to examine emerging questions about technology, ethics, and sentient life in a rapidly changing world. Join us from Friday 30th May until Sunday 1st June to work together to create a brighter future for all sentient beings. We have both in-person and virtual options across all three days.
Friday: A classic conference day with talks, panels, and workshops delivered by leaders thinking about the impact of AI on sentient non-humans – now and in the future.
Saturday & Sunday: A smaller unconference – where you set the agenda! Pitch topics, lead discussions, and dive into what matters most. We have found that this interactive format sparks fresh ideas and leads to more serendipitous connections.
Early bird discount until the end of April with code: EB25AIADM
We’d love to hear your questions or feedback! Email hello@aiforanimals.org.
You can subscribe to this newsletter here.
Index
Conference Retrospective
We recently wrapped up AIA Bay Area 2025, bringing together over 500 hybrid participants for two days of collaborative learning and community building. We have compiled some highlights below, but you can also…
Beyond the main conference, participants took part in a week filled with professional and social satellite events.
With 80+ unique sessions ranging from formal presentations by speakers like Peter Singer and Jeff Sebo to community-led unconference sessions, attendees were able to customize their own experience.
Many expressed appreciation for the rare opportunity to engage deeply with topics at the intersection of artificial intelligence and nonhuman welfare, including interspecies communication, welfare monitoring, digital minds, and conservation technology. See the testimonials here.
Special thanks to Constance Li, Sankalpa Ghose, and Santeri Tani for bringing the retrospective to life with their thoughtful reflections and insights.
🚨Updates
Community Poll
We want to make this newsletter as useful as possible. Please take 1 minute to fill out the poll below, or email us at hello@aiforanimals.org with any ideas and feedback!
New YouTube Playlists
We have created new playlists to help our YouTube subscribers navigate our content:
We hosted a fireside chat with Robin Larson from Anthropic and discussion groups on transformative AI (TAI) on Sunday, April 6th.
View Robin’s recap on AI’s potential to help or harm animals posted in our Slack community here.
Did you know that the Bay Area has one of the highest densities of animal advocates globally, yet most work remotely without regular in-person collaboration? We are working to change that!
Last Thursday, April 10th, we provided a space at Mox for our first pro-animal open coworking session.
Now that we have identified demand for a collaborative space in the city, we want to move forward with creating an innovation hub for animal advocates in SF. You can learn more about and support our project on Manifund.
📅 Upcoming Events
Our Events
Cross-Movement Learning: Animal Welfare Strategies for AI Safety Advocates
Tuesday, April 29 at 6:30 pm PT in San Francisco, California
You won’t want to miss this fireside chat with Leah Garcés, CEO and President of Mercy for Animals, as she brings her passion and insight to a conversation on advocacy strategies for the AI safety community.
Find us at…
Thursday to Sunday, May 15-18 in Los Angeles, California
At this event connecting 800+ global farmed animal advocates, our Founder and Executive Director, Constance Li, will be moderating the Opening Ceremony and speaking on the panel: Field-building at the Intersection of AI and Animal Welfare.
Friday to Sunday, June 6-8 in London, England
One week after AIADM 2025, join us at the annual London EAG conference where we will (co-)host a few events including pro-animal coworking on Friday and an animal welfare meetup on Saturday.
Friday to Sunday, October 10-12 in New York City
Exciting things are taking shape this fall as we plan satellite events to grow the AIxAnimals community on the East Coast—stay tuned!
Other Events
Prospects and Pitfalls for Real Artificial Consciousness
Wednesday, April 16 at 1:00 pm ET (virtual)
Anil Seth examines the science of consciousness in this talk hosted by the NYU Center for Mind, Ethics, and Policy.
Thursday to Sunday, August 21-25 in Berlin, Germany
There are 40 spots available for ambitious changemakers from across the world for this summit hosted by the Compassionate Future Initiative.
🌏 Opportunities
Submit a paper to one of the following journal editions at the intersections of AIxAnimals
Ostrich – Journal of African Ornithology - Integration of Artificial Intelligence (AI) in Ornithology - Submit by June 15th
The Applications of Emerging Technologies on Biodiversity Conservation - Submit by June 30th
Topoi (Philosophy) - Interactions between Humans, Non-human Animals, and AI: Philosophical Underpinnings and Ethical Considerations- Submit by July 31st
Animals - Artificial Intelligence as a Useful Tool in Behavioural Studies - Submit by October 31st
🗞️ News & Research
🗣️ Understanding animals
Animals in Translation: Imagining Criteria and Frameworks for Decoding Communication in Other Species (Interspecies Internet)
This publication shares the outcomes and proceedings of the 2024 Interspecies Internet ‘Animals in Translation’ Workshop, representing a significant advancement in our understanding of animal communication systems and establishing new standards for the field.
AI systems are now outperforming humans at detecting pain and stress in animals by analyzing facial expressions, enabling more individualized and timely care on farms and in shelters. Researchers aim to expand this to complex emotions like happiness and frustration, raising both welfare opportunities and ethical challenges around interpretation and deployment.
Humpback whale song shown to be structurally similar to human language (Phys.org)
Researchers have found that humpback whale songs share one of the same statistical structures found in all human languages, suggesting deep parallels in how complex communication systems evolve culturally. This challenges assumptions about human uniqueness and could imply that elements of language-like learning are widespread across species.
🐔 Chicken farming
AI is for the birds: How machine learning can help predict and manage avian flu outbreaks (The Conversation)
AI models can predict avian flu outbreaks up to four weeks in advance by analyzing diverse data sources such as wild bird reports, climate data, farm records, and online media, achieving 85% accuracy in trials. These tools enable early interventions, helping prevent economic losses, food insecurity, and potential human pandemics, but require strong cross-sector partnerships and responsible data governance to be effective.
Marel introduces artificial intelligence in poultry processing (Poultry World)
Marel’s new ‘vent-cutting’ machine uses AI to cut broiler carcasses more precisely and consistently than human operators, helping prevent contamination and reduce waste. It adjusts settings automatically, tracks performance in real-time, saves water when no birds are present, and works at very high processing speeds.
🐛 Animal farming: General
Polish biotech firm unveils scalable beef analog from insects (PetFoodIndustry.com)
Proteine Resources has created an insect-based beef substitute for pet food that matches real beef’s nutrition, cuts production costs by up to 50%, and may reduce pet allergies. It’s scaling up with AI-powered mini-factories and plans to sign major supply deals by 2026.
Food Standards Agency brings AI to meat inspection (Public Technology)
The UK Food Standards Agency is piloting AI tools to support meat inspectors by using mobile apps that streamline note-taking and improve data consistency. It is also applying both generative and traditional AI to clean and categorize unstructured food safety data and detect risks in documents like shipping manifests before products enter the country.
🦘Wild animals
Harnessing artificial intelligence to fill global shortfalls in biodiversity knowledge (Nature)
Artificial intelligence (AI) has the potential to address seven key gaps in biodiversity knowledge by integrating diverse data types—such as images, audio, and DNA—to enhance ecological research. Implementing AI in these areas could significantly advance our understanding of biodiversity and aid in achieving the 2030 targets of the Global Biodiversity Framework.
3 new ways we’re working to protect and restore nature using AI (Google)
Google has launched three AI initiatives to protect nature: a startup accelerator for tech solutions in conservation; a $3M grant fund for Brazilian nonprofits using AI in biodiversity, agriculture, and bioeconomy; and the open-source release of SpeciesNet, an AI model for wildlife photo identification.
🍔 Alternative proteins
AI Helps To Create Lab-Grown Meat (Forbes)
AI is helping scientists grow meat in labs more efficiently by testing which proteins and growth signals—molecules that tell cells when to grow and divide—work best. For example, AI identified improved versions of these signals that made cells grow faster, helping companies like GOOD Meat and Upside Foods produce more cultured meat with less trial and error.
Pow.Bio Opens New Facility With AI-Enabled Continuous Fermentation Capabilities (Vegconomist)
Pow.Bio has opened a new 25,000 sq ft facility in California featuring AI-enabled continuous fermentation, enabling rapid, scalable production of bio-based ingredients with reduced costs and development time. The facility, equipped for food-grade manufacturing, supports partners like MeliBio and claims up to 5× productivity versus traditional fermentation, setting a new industry benchmark.
NotCo’s AI is Revolutionising the Food Industry – And Big Brands Are Eating It Up (Green Queen)
NotCo uses its proprietary AI platform, Giuseppe, to develop and reformulate food products by optimizing ingredient combinations for taste, nutrition, and scalability—reducing R&D timelines from years to months. Now a B2B-focused tech company, NotCo works with major corporations like Kraft Heinz, Nestlé, and Mars to improve formulations, launch new plant-based products, and integrate its AI into their R&D pipelines.
🌱 Crop farming
Google’s X spins out Heritable Agriculture, a startup using AI to improve crop yield (TechCrunch)
Heritable Agriculture, a Google X spinout, uses AI to analyze plant genomes and identify optimal breeding combinations to improve crop yields, reduce water use, and boost carbon storage—without relying on genetic modification. The startup tested its models on thousands of plants in labs and field sites, and is now focused on commercialization with support from investors including Google.
Is Artificial Intelligence the future of farming? Exploring opportunities and challenges in Sub-Saharan Africa (World Bank)
AI is transforming agriculture in Sub-Saharan Africa through tools like precision farming, pest detection, and predictive analytics, improving yields and sustainability while enabling services such as tractor-sharing, crop disease diagnosis, and market access. However, adoption faces barriers including limited connectivity, high costs, lack of digital skills, and weak data governance, requiring coordinated policy action and infrastructure investment.
🖥️ Alignment
AI Moral Alignment: The Most Important Goal of Our Generation (Ronen Bar, EA Forum)
Moral Alignment aims to guide AI toward valuing all sentient beings, not just humans, to avoid repeating ethical failures. Despite its urgency, it's severely underfunded and needs immediate, coordinated attention.
What do Large Language Models Say About Animals? Investigating Risks of Animal Harm in Generated Text (Arturs Kanepajs et al., Arxiv)
This study introduces the Animal Harm Assessment (AHA), a novel method for evaluating how large language models might promote or reduce harm to animals through their outputs. Applying AHA to frontier LLMs reveals substantial variation in how different models handle ethically relevant animal scenarios, highlighting the urgent need for animal-specific benchmarks in AI safety.
🐁 Animal testing
AI, MPS and the End of Animal Testing? (Technology Networks)
AI and microphysiological systems (MPS) are reshaping drug testing by offering more human-relevant, efficient, and ethical alternatives to animal models. AI enhances predictive toxicology and data integration, while MPS like organs-on-chips mimic human organ function, improving early detection of drug safety issues—though regulatory acceptance remains a key barrier to full replacement of animal testing.
AI-Powered Tech Enables Continuous Lab Animal Monitoring (The Scientist)
AI-powered home-cage monitoring systems now let researchers track lab animals continuously and in detail, supposedly improving behavioral data quality, reproducibility, and animal welfare. These systems, like DAX and smart-Kage, use machine learning to detect complex behaviors with minimal human interference, enabling earlier health alerts, reduced stress, and fewer animals needed for research.
🐶 Companion animals
AI robot pets can be adorable and emotionally responsive. They also raise questions about attachment and mental health (The Conversation)
AI robot pets like Ropet can offer emotionally responsive companionship, but raise concerns about unhealthy attachment, especially among children and vulnerable individuals. These devices also pose privacy and security risks due to their data collection and cloud-based processing, underscoring the need for regulation and safeguards before they become widespread.
🤖 Digital minds
We’re Thinking About Deploying AI Models Which Have An “I Quit” Button: Anthropic CEO (OfficeChai)
Anthropic CEO Dario Amodei has proposed giving AI models an “I quit” button, allowing them to refuse tasks—an idea rooted in the possibility that advanced models might have meaningful subjective experiences. While speculative, the proposal signals growing concern over AI welfare and introduces practical mechanisms to detect and respect potential model preferences.
🚀 …and more
How should we adapt animal advocacy to near-term AGI? (Max Taylor, EA Forum)
If AGI arrives within 5–10 years, it could either massively reduce or vastly amplify animal suffering depending on whose values it reflects. Animal advocates must act now to shape AGI’s moral foundations; waiting risks value lock-in that entrenches exploitation at unprecedented scale.
That’s it for this edition - as always, please feel free to get in touch at hello@aiforanimals.org with any ideas and feedback!
https://www.fda.gov/news-events/press-announcements/fda-announces-plan-phase-out-animal-testing-requirement-monoclonal-antibodies-and-other-drugs
Thanks Max - looking forward to London. My recent conversation with Ronen Bar for Sentientism might be of interest to your readers to build on his EA Forum post you reference: https://youtu.be/9hDIQj-i44M