Advancements in artificial intelligence are unlocking the complexities of animal languages, prompting researchers to introduce a structured ethical framework to guide these efforts responsibly.
Breakthroughs Meet Hidden Dangers
Breakthroughs Meet Hidden Dangers (Image Credits: Wikimedia)
Scientists at New York University’s More than Human Life Program (MOTH) and the Cetacean Translation Initiative (CETI) published the PEPP Framework in November 2025 as part of their report, “Listening to Our Animal Kin.” This initiative emerged from interdisciplinary workshops and addresses the rapid rise of nonhuman animal communication technologies, or NACTs, which employ AI and machine learning to record, analyze, and interpret signals from species like whales, elephants, and honeybees.
These tools promise conservation gains, such as decoding whale distress calls triggered by shipping noise to push for stronger protections. Yet, early experiments revealed unintended consequences. In one case, researchers played a recording of a deceased elephant’s call, sparking chaos in the family group as they searched frantically for the missing member; the daughter continued calling for days. CETI founder David Gruber noted, “Even routine recording and playback can cause stress in animals.”
The PEPP Pillars in Action
The framework organizes its guidance into four pillars – Prepare, Engage, Prevent, and Protect – each supported by three principles, totaling 12 actionable standards. Under Prepare, researchers must establish rigorous protocols incorporating the 3Rs: replace animal involvement where possible, reduce numbers affected, and refine methods to limit harm. This includes consulting veterinarians and publishing transparent ethics and data governance plans before deployment.
Engage calls for stakeholder input from species experts, Indigenous communities, and conservationists, ensuring free, prior, and informed consent where applicable. Prevent shifts the burden of proof to justify risks, mandating comprehensive analyses of physical, mental, relational, and ecological impacts, with a precautionary halt if serious threats emerge. Protect prioritizes animal autonomy, best interests, legal compliance, and harm remediation.
- Prepare: Rigorous research design and governing protocols.
- Engage: Diverse consultations, recognition of contributions, and transparency.
- Prevent: Justification of risks, mitigation strategies, and precaution.
- Protect: Autonomy respect, best interests assessments, implementation diligence, and remediation.
Risks Beyond the Lab
NACTs carry broader perils, including data misuse for poaching via shared locations or military applications to manipulate wildlife. Mental distress from surveillance or fake signals, relational disruptions in social groups, and ecological ripple effects like acoustic pollution further underscore the need for safeguards. The framework draws from established standards, such as the EU’s animal welfare directive and AI Act, adapting human rights analogies like the “best interests of the child” to nonhuman subjects.
Historical precedents offer hope. The 1970s revelation of humpback whale songs spurred legal protections for the species, demonstrating how ethical insights can drive policy. Gruber emphasized that voluntary principles like PEPP could evolve into binding international norms, much like human rights standards.
Toward Responsible Innovation
Co-authored by MOTH’s César Rodríguez-Garavito, Jacqueline Gallant, and Emma Crowe, the framework positions ethics as foundational to innovation, fostering kinship over control. It urges funders, engineers, and policymakers to adopt these living principles, which remain open to iteration based on new evidence.
By centering nonhuman dignity, PEPP ensures AI amplifies empathy and coexistence rather than exploitation. What role should society play in shaping these interspecies conversations? Share your thoughts in the comments.
Key Takeaways
- PEPP provides 12 principles across four pillars to govern AI animal communication tech.
- Real-world harms, like elephant distress, highlight the framework’s urgency.
- Voluntary now, it paves the way for enforceable global standards.
