Many thanks to Alina Salmen, Vince Mak, Constance Li, and Johannes Pichler for feedback on this post. All mistakes are our own. This post does not necessarily reflect the views of our employers.

Introduction

Rapid AI development presents unprecedented opportunities and significant challenges for animal advocacy. AI could either worsen animal suffering by, e.g, making exploitative systems more efficient, or drastically reduce it by enabling new and improving current solutions. The stakes are immense: AI could profoundly influence the trajectory of animal welfare at a scale we have not seen before - and it could go in either direction. Understanding these potential shifts now is crucial for developing proactive strategies and ensuring our movement's long-term effectiveness.

This piece explores the evolving roles of existing animal advocacy interventions in a post-AI society, looking at how they may change in their nature, feasibility and cost-effectiveness. We don't attempt to assess the likelihood of any particular intervention being affected in any particular way; instead, we hope to pose questions that we believe to be underexplored and important if we take the possibility of transformative AI seriously.

We explore five different interventions in depth in this post (Corporate and Institutional Outreach for Welfare Improvements, Network Building, Government Outreach, Corporate and Institutional Veg*n Outreach, and Research). We have also drafted a list of the potential effects on all 27 interventions listed in Animal Charity Evaluators’ Menu of Interventions, and will keep this updated. If you’d like access to this list, please just fill out this Google Form. We welcome collaborative input as we continue to refine these ideas.

Common Patterns and Broader Implications

Before delving into specific interventions, we've identified several overarching patterns that consistently appear across interventions.

• Symmetric AI Access and Application: AI's use cases are largely symmetrical: the tools that empower animal advocacy are equally accessible to the animal agriculture industry and counter-movements. Keeping this symmetric nature in mind when approaching interventions is crucial. The challenge lies in capitalising on potentially beneficial asymmetries: for example, perhaps newer alternative protein startups and advocacy organizations can more easily pivot to incorporate new AI capabilities than entrenched animal agriculture systems with complex, interdependent systems consisting of inflexible physical infrastructure.
• Predictive Modeling and Strategic AI Advisory Systems: Predictive AI can identify promising targets, asks, and strategies for a variety of advocacy interventions, acting as a multi-purpose tool. This could unblock interventions that are currently bottlenecked by identifying the right target, ask or strategy. Ensuring that we have the data infrastructure necessary to leverage this application becomes crucial.
• Automation and Shifting Labor Bottlenecks: AI could automate many tasks, particularly in knowledge work. This could level the playing field for under-resourced groups by making high-quality services such as writing literature reviews, writing grant proposals, and drafting basic legal text more accessible. While roles like these may become less important from  automation, those relying on effective AI utilization, human connection and trust-building may become more important.
• The Attention Landscape: Overwhelm, Authenticity, and Filter Bubbles: The sheer volume of AI-generated content poses challenges. The media landscape could be flooded, making it harder for any content to stand out. "Deepfakes" threaten the credibility of visual evidence, impacting investigations and endorsements. AI-driven personalization and segmentation could intensify "filter bubbles," making it difficult to reach those not already receptive to pro-animal messages. New, unconventional ways to reach people might become essential.

Our movement’s preparedness for these trends will determine if AI becomes a force multiplier or a significant impediment for animal advocacy. Among other things, this broadly implies the following for our work:

• Animal advocates must remain highly flexible and adaptable, continuously re-evaluating strategies and tactics in light of rapid AI developments. What works today may be obsolete tomorrow, or conversely, a new, highly effective avenue may emerge unexpectedly.
• In order to help prioritise our work today, we should consider which interventions may become significantly more or less cost-effective in the future, which may become riskier, and which tasks AI may soon be able to automate.
• It is crucial for the animal advocacy movement to diligently track how the animal agriculture industry and other oppositions are leveraging AI. Understanding their advancements and applications can help prevent our efforts from being blindsided and inform our own defensive and offensive strategies.
• In order to seize opportunities and mitigate risks, establishing the necessary infrastructure—such as data availability/pipelines, AI tools, and specialized expertise—and developing agile strategies to navigate emerging bottlenecks, may prove crucial and should be implemented sooner rather than later. Animal advocates should share best practices for effective AI use with the community to foster collective upskilling.

We believe that strategically investing in AI capacity, fostering a culture of continuous learning, and maintaining a vigilant eye on the evolving landscape are increasingly vital for ensuring AI benefits animals. We need to act now to seize this narrow window to establish a decisive, relative advantage for animals amidst profound technological shifts.

Deep Dive into Key Interventions

We've chosen to explore five key interventions in greater detail—Corporate and Institutional Outreach for Welfare Improvements, Network Building, Government Outreach, Corporate and Institutional Veg*n Outreach, and Research. They are among the most heavily funded interventions and also offer rich ground for illustrating the complex opportunities and challenges posed by AI.

You can access the full draft document we created exploring the effects on all interventions as categorized by ACE’s Menu of Interventions, by filling out this Google Form.^[1]

Corporate and Institutional Outreach for Welfare Improvements

Brief Explainer: Persuading companies and institutions to commit to improving animal welfare standards through their supply chains.

• Opportunities:
  • Optimized Targeting & Strategy: Predictive modeling may help identify particularly promising companies, optimal 'asks,' and most effective communication strategies for each stakeholder.
  • Enhanced Accountability: In countries with progressive legislation or corporate transparency, NGOs and/or government agencies could gain continuous access to farm footage. AI could then analyze this footage to ensure corporations uphold welfare commitments, becoming increasingly effective as farms adopt Precision Livestock Farming (PLF) methods. AI could also help charities gather and interpret vastly more relevant material for accountability, such as sifting through supply chain records, public reports, and satellite imagery, leading to drastically more and better transparency mechanisms.
  • Shifted Welfare Standards: The standard 'asks' for corporate welfare campaigns could evolve. Instead of current approximate welfare improvements like 'going cage-free,' NGOs could demand robust AI use by companies to ensure verifiable and measurable welfare outcome improvements on farms.
• Risks/Obstacles:
  • Corporate Counter-Optimization: AI could assist corporate decision-makers in simulating potential outcomes of policy changes or campaign responses. This could be good or bad for animal welfare, depending on whether decision-makers relatively over- or underestimate the impact of public pressure campaigns on their economic bottom line. There is also a symmetrical risk; just as advocates can use AI to improve their communications strategies, so too will corporations have drastically improved messaging and marketing.
  • Gaming the System: Companies might increasingly argue that advanced PLF renders familiar infrastructure changes, like going cage-free, unnecessary. If done disingenuously, this could take the form of companies 'gaming the system' to make animals' welfare seem better than it is, or even arguing that factory farming, with AI, is a force for good.
  • Undermined Theories of Change: Even when PLF genuinely improves welfare, it may also boost productivity and profitability. This complicates long-term theories of change, as animal agriculture could become more economically sustainable just as it becomes more welfare-optimized. TAI could thus undermine the strategy of making animal agriculture less profitable through incremental welfare demands.
• Unclear Effects & Considerations:
  • Reliance on Broader Interventions: Corporate campaigns often tie into social media, protests or other interventions to help escalate campaigns. Their effectiveness post-TAI would depend on how these supplemental interventions are impacted by AI. It is currently unclear whether post-TAI corporate campaigning would largely tie to a different, more effective supplemental intervention (such as AI-aided auditing).

Network Building

Brief Explainer: Creating and strengthening connections, alliances, and coalitions within animal advocacy and with other movements.

• Opportunities:
  • Enhanced Human-Led Coordination: AI could help develop movement-level theories of change and dynamically allocate roles across groups to ensure complementary strategies and avoid duplication. This currently requires significant manual input and negotiation.
  • Strategic Collaboration Identification: AI could facilitate targeted coordination by using data on network members to recommend meetings most likely to generate significant impact. Predictive modeling could identify key connectors or potential collaborators across movements.
  • Removing Language Barriers: AI could facilitate wider global coordination through seamless translation and interpretation.
  • Leveraging Pooled Knowledge: Knowledge sharing could become more useful as AI models are able to use pooled knowledge from our networks to come up with sophisticated insights and recommendations, overcoming the current challenge of information overload for most advocates.
  • Conflict Resolution: AI tools like the ‘Habermas Machine’ could find common ground between differing factions within a movement, reducing the likelihood of infighting.
• Risks/Obstacles:
  • Reduced Importance of Human Advocates: Human advocates may become less important if AI agents can autonomously plan campaigns and generate outreach materials, potentially reducing the perceived need for traditional human networks.
  • New Sources of Infighting: Conversely, AI-driven changes could give rise to new sources of friction within the movement (such as disagreements about how much to prioritise new forms of digital advocacy over traditional human-focused methods) that increase infighting if not managed carefully.

Government Outreach

Brief Explainer: Engaging with politicians and government institutions to change laws and policies affecting animals.

• Opportunities:
  • Optimized Policy Targeting: Predictive modeling of legislative outcomes could improve strategic focus on winnable policy changes. Advisory AI systems could suggest the most promising politicians to engage with and the most compelling communication strategies to use.
  • Mass Citizen Advocacy: AI could facilitate mass government outreach by the public. Advisory AI systems could suggest promising demographics to target, and NGOs could then encourage citizens to use AI to more easily craft personalized, compelling messages to send to their representatives or other political decision-makers.
  • Streamlined Legislative Drafting: AI legislative drafting assistance might enable advocates to produce more sophisticated policy proposals with fewer specialized legal resources.
  • Finding Common Ground: AI tools like the ‘Habermas Machine’ mentioned above could find common ground between advocates and policy-makers, helping them see eye-to-eye and facilitating mutually beneficial outcomes.
  • Increased Air Time for Frontier Issues: AI-driven efficiencies could free up policy-makers to deal with more issues than they currently have time for,  including issues like animal welfare, which are currently overlooked.
• Risks/Obstacles:
  • Institutional Agility Gap: If political institutions are less agile during accelerated AI development (e.g., due to bureaucracy and risk aversion), their role might be limited to "not blocking" pro-animal efforts (e.g., alternative protein development) rather than actively advancing them. Advocacy might be put to better use if targeted towards institutions such as corporations that are able to make more effective use of AI and therefore come to hold more power.
  • Symmetrical Opposition Power: Many of the same AI tools that advocates can use to optimise outreach will also be accessible to the opposition, which often has significantly greater financial and human resources. As a result, advocates may face stronger, more sophisticated opposition in the policy space than ever before.
  • Overwhelm of AI-Generated Messages: If many organizations use AI for mass citizen advocacy and it leads to a surge in formulaic, AI-generated messages, governments may become overwhelmed or dismissive. This could render the tactic ineffective or even reputationally damaging if perceived as manipulative or inauthentic.
  • Bottlenecks of Human Networks: Policy advocacy may largely remain bottlenecked by personal networks and trust, which are more reliant on slow, interpersonal relationship-building that cannot easily be automated, and where animal advocacy is generally at a disadvantage compared to, e.g., the animal agriculture industry.
• Unclear Effects & Considerations:
  • Governmental Control of AI: If AI systems come under governmental control (e.g., through nationalization or strict licensing), they may become more powerful and influential than ever, making policy leverage even more critical. In this scenario, shaping AI-related regulation could become especially high-stakes, as government-owned AI systems might strongly influence public policy, industry norms, and resource allocation.

Corporate and Institutional Veg*n Outreach

Brief Explainer: Transforming food environments in companies and institutions to promote plant-based options and reduce animal product consumption.

• Opportunities:
  • Optimized Targeting & Strategy: Predictive modeling may enable advocates to identify particularly promising companies and institutions, along with effective 'asks' and communication strategies. More generally, TAI could vastly improve our understanding of the behavioral science behind dietary choices through large-scale data collection, analysis, and simulation studies, enhancing the design and targeting of veg*n outreach strategies across the board.
  • Better Alternatives: If AI accelerates the development of alternative proteins, we may have more attractive options for corporations, institutions, and their stakeholders. This highlights a case for scaling up veg*n outreach now, before TAI, specifically to build experience, test interventions, and establish institutional infrastructure, making future outreach more cost-effective once taste and price parity—or superiority— are achieved. It also highlights removing potential obstacles in adopting alternative proteins, such as political bans or narratives about ultra-processed foods being unhealthy.
  • Alignment with Institutional Priorities: Institutions may be able to accurately predict how specific menu changes will help them achieve their desired goals (e.g., climate goals, employee/student well-being, reputational benefits). This could be beneficial if advocates successfully align animal-friendly outcomes with these existing institutional priorities.
• Risks/Obstacles:
  • Deprioritization of Animal Welfare: If AI systems give institutions precise control over process optimisation, they may find ways to fulfil top-priority goals like cost, health, or sustainability without benefiting animal welfare, making advocacy more challenging. This could even worsen animal welfare if, for example, institutions make a sustainability-driven shift from beef to fish.
• Unclear Effects & Considerations:
  • Impact on Nudges: In an AI-transformed world, subtle promotional strategies like nudges may become less useful (if people are better empowered to make autonomous decisions or rely on AI agents) or more useful (if people are overwhelmed by choice). Traditional nudges could become less effective unless embedded in AI’s architecture, or conversely, more powerful as tools to simplify choices amidst information overload.

Research - Effective Advocacy, Farmed Animal Welfare Science, and Wild Animal Welfare

Brief Explainer: Investigating strategies, interventions, and cause areas to identify the most impactful ways to help animals and advance scientific understanding of animal welfare.

• Opportunities:
  • Accelerated Research Capacity: From a capacity perspective, AI may drastically benefit research, especially areas requiring fewer real-world, context-rich, long-time scale experiments. Even for complex experiments, AI could largely resolve bottlenecks by enabling massive simulations.
  • Improved Resource Allocation: Improved research could allow for a far more effective allocation of resources; for example, by identifying animals’ capacity to feel pleasure and pain (like a far more advanced version of current helpful projects like the Moral Weight Project). AI research tools and modeling could provide a better sense of promising advocacy interventions and organizational tactics.
  • Converging on "True" Answers: AI-driven research on wild and farmed animal welfare may point to whatever the "true" answer is regarding animal sentience, suffering, and flourishing, which seems to favor animal advocates.
  • Accelerated WAW Modeling: Transformative AI could greatly accelerate theoretical and modeling work on wild animal welfare. This suggests it might be more effective to focus on experimental, physiological fieldwork that AI will take longer to fully automate. (See Transformative AI and wild animals: An exploration for further discussion on this point.)
  • Interspecies Communication: Some organizations are exploring AI to enable interspecies communication. In principle, this could expand people’s moral circles and increase the likelihood of according basic legal rights to other species.
• Risks/Obstacles:
  • Symmetrical Research for Counter-Movements: Research on effective animal advocacy, in particular, may be very symmetrical, as counter-movements can similarly research effective ways to stop animal advocacy.
  • Credibility Risks from Overreliance: Overreliance on AI-directed research without sufficient human oversight and quality checks could lead to erroneous conclusions and damage advocates’ credibility.
  • Frivolous Use of Interspecies Communication: The impact of interspecies communication depends on the expertise and motivations of developers and the public narrative formed. It is possible that such technology could be used for frivolous human entertainment, with little regard for fundamental ethical implications.
• Unclear Effects & Considerations:
  • Shifting Bottlenecks (Management): Research prioritization judgment, agenda-setting, and management may become a more important bottleneck if these areas are less automatable, although even these may become less critical with advanced simulations.

Conclusion

The journey into an AI-transformed world is one of profound uncertainty, but also immense potential. For animal advocacy, this means navigating a future where AI could reshape every aspect of our work, from how we gather evidence to how we influence public opinion and policy. By proactively understanding the common patterns—the symmetrical access to AI's power, the rise of predictive tools, the shifts in labor bottlenecks, and the evolving information landscape—and by strategically investing in the right infrastructure and skills, we can aim to harness this transformative technology to accelerate progress for animals. The time for strategic engagement with AI’s potential for animal welfare is not in the future; it is definitely now.

Appendix: Other Animal Advocacy Interventions

We have also drafted a list of the potential effects on all 27 interventions listed in Animal Charity Evaluators’ Menu of Interventions, and will keep this updated. We don’t want to make any claim on exhaustiveness, prioritization, or likelihood of any particular effect. The document is a continuous work in progress and should primarily act as a nudge and inspiration to dive deeper into any particular intervention and their respective opportunities and obstacles, rather than an exhaustive overview of the possible effects.  If you’d like access to this list, please just fill out this Google Form. We welcome collaborative input as we continue to refine these ideas.

1. ^{^}

    Given the expansive potential impacts of AI across the animal advocacy landscape, a comprehensive deep dive into every intervention would be unwieldy for this post.