Summary from Food Safety Coalition Workshop, May 2021

UN Food Systems Summit. 05/14/2021 - 05/14/2021

On May 14, 2021, the first Food Safety Coalition meeting took place, bringing together a group of like-minded organizations to collaborate and progress solutions in the critical area of food safety, specifically data and knowledge sharing to address the challenge of aflatoxin contamination in raw materials.

Contact Us

On May 14, 2021, the first Food Safety Coalition meeting took place, bringing together a group of like-minded organizations to collaborate and progress solutions in the critical area of food safety, specifically data and knowledge sharing to address the challenge of aflatoxin contamination in raw materials. The aim of this coalition is to identify specific actions to share in the run up to the UN Food Systems Summit in September 2021, and to keep working on them beyond the summit to progress a long-term transformational food safety agenda. As discussed in the CEO Consultations for the UN Food Systems Summit, we all have a role to play in making food systems better through safe food – all of us includes the non-governmental organizations, regulators, academia, and the private sector.

The following participated in the first workshop of the Food Safety Coalition:





Dr. Amare Ayalew

Partnership for Aflatoxin Control in Africa (PACA)


Bob Baker



Cornelia Boesch

The United Nations Food and Agriculture Organization (FAO)


Chuck Bird

Neogen Corporation


Dr. Susan Blount



Dr. Christopher Elliott

Queen’s University Belfast


Jean-Christophe Flatin



Dr. Christopher Gilligan

The University of Cambridge


Dr. Jagger Harvey

Kansas State University


Yi Fan Jiang

Food Industry Asia


Matthew Kovac

Food Industry Asia


Dr. Markus Lipp

The United Nations Food and Agriculture Organization (FAO)


John Luedke



T.W. Lim

Gro Intelligence


Mary Mureithi

The World Food Programme


Jared Shaw



Virginia Siebenrok

The World Food Programme


Dr. Abigail Stevenson



Dr. Maria Velissariou



Dr. Guangtao Zhang


There are many food safety challenges that must be addressed and the Food Safety Coalition is focusing first on aflatoxins given the serious health threat they pose:

  • Aflatoxins are a key global challenge, posing a serious threat, in the developing and developed world. As well as being linked to stunting, damage to the immune system, and maternal anemia, they are estimated to play a part in 28% of liver cancers globally. Evidence also suggests aflatoxins are a challenge in mature economies and could become more so because of climate change. In developed economies, when aflatoxin levels are found to be above legal limits, food is disposed of, leading to food waste and an environmental risk and health risk as decontamination is not easy.
  • Current solutions to tackle aflatoxins are imperfect and must be improved. Aflatoxin control can be targeted to both pre-and post-harvest interventions. Pre-harvest primarily focuses on managing the extent of toxigenic spore contamination; while post-harvest is focused on minimizing toxin formation; both are largely dependent on the application of good agricultural practice. Mitigation of aflatoxin is typically conducted through physical, chemical, and biological means. Lack of effective decontamination methods for aflatoxin tainted crops remains a major gap and has led to significant food waste and food fraud, and there is an additional, significant challenge of a lack of globally harmonized regulatory limits for aflatoxin.
  • The application of an aflatoxin prediction model with the free exchange of data, as well as putting data into context in this space, can enable the right conversations and interventions. Mars has built a prediction model in collaboration with the University of Cambridge, linking the level of aflatoxin with climatic conditions and intervention measures.
  • This is an area that requires a joint effort and a pre-competitive solution. We can help identify the data needed for collection and analysis to make a difference, and the effort would require others—from businesses to governments to international organizations—to share their insights, develop the platform, and contribute the data.
Food Safety Coalition Workshop: Key Themes

During the webinar, participants discussed the impacts and challenges of aflatoxin risk management, the current tools available and lessons learned from early application, as well as the key challenges surrounding prediction models and adoption of advantageous practices more broadly. Key themes from the discussion are captured below.

1. We are not starting from scratch in generation of aflatoxin risk management knowledge and although there are gaps, there are existing, ongoing programs that we can learn from to unlock and drive new solutions

  • It is important to look at existing initiatives, for example the Partnership for Aflatoxin Control for Africa (PACA) work seeking to bring existing efforts together across Africa, leverage them and learn from them. There is also interest in having a similar shared platform for data generation and analytics in Asia that would allow linking up of datasets, and early adoption of common sampling protocols to ensure consistency of data collection. 
  • There are many programs such as the Post-Harvest Loss Initiative between USAID and Kansas State University trying to remove aflatoxin, and insights from these could help refine the aflatoxin model.
    1. There are data from projects in Ireland, led by the Queen’s University, Belfast that could also inform development of solutions.

2. Sampling and testing are central to finding sustainable solutions in aflatoxin risk management, yet they are also the biggest challenges

  • The quality and robustness of data that goes into sampling and testing is key and will impact accuracy of a prediction model and prediction capabilities. The harmonization of sampling and testing protocols is also important. Drivers for change include both regulatory compliance and supplier quality audits.
  • There is a need to improve methodology to ensure an accurate assessment of the burden of contamination.
  • Rapid methods can be very powerful, allowing in-field analysis and factory gate sampling. To be successful these methods require a four-pillar approach: accuracy; ease of use; timely generation of results (allows for running multiple samples); and affordability. The priority order of these four factors may vary depending on the priorities or needs of the local area.
  • High quality data input correctly can also enable helpful decision support tools. For example, by using a smart phone to collect disparate data, uploading them for interpretation and then feeding them back to those across the supply chain, it’s possible to advise and support decision making on the ground in a way that’s easy to understand.
  • Infrastructure is a consideration to help people reliably obtain and access data (for example, Wi-Fi access is not universal).
  • There is a capability building need: many practitioners don't understand the importance of proper sampling (even some researchers conducting surveys). There is an opportunity to map out what successful programs are already doing in capability building and build a support platform to help guide sampling and national efforts to generate data. There is scope to approach other organizations such as Scientific Animations Without Borders[1] to help guide sampling. 

3. Data sharing – opportunities and concerns

  • There is a need to create a safe space for people to share data anonymously – (data is required to help strengthen insights and preventative efforts in food safety but many are concerned that without anonymity, there could be an impact on their business and livelihoods).
  • Additional Data generation support is still needed in parts of Africa.
  • A harmonized approach (for example, through organizations such as PACA for Africa) could also help to reduce unintended impacts on any one individual providing data.
  • Coalition efforts could potentially leverage data from GEMS/Food (Global Environment Monitoring System - Food Contamination Monitoring and Assessment Programme)[2]
  • An approach to risk and education needs to be understood to evaluate challenges, opportunities, and root causes (and ensure understanding of how data can be useful as well as detrimental).
  • There are opportunities for fraudulent activity if individuals feel threatened by the consequences of data generation and sharing (sampling and testing can drive non-food safe behaviors such as falsification of data if there are uncertainties on how the outcomes and actions can impact business performance). 
  • It’s important to consider unintended negative consequences - once people have access to robust data indicating areas with high levels of contamination, they could choose not to source from certain areas. Example: because of awareness of high level of aflatoxins in a certain known region, small holder farmers may not be able to sell their crops and no longer afford the cost of interventions.

4. Proper handling of contaminated materials is also a challenge

  • Scaling can be paired with markets; meaning, there is potential opportunity to reduce costs for smallholder farmers, who are already impoverished in many cases.
  • It’s important to consider what is done with rejected material (there is a risk that food contaminated with aflatoxin can be consumed by most vulnerable members of society).
  • There are potential alternative uses for rejected material OR decontamination is a possibility which must be part of a mitigation plan.
  • Mitigation methods should include what we do when our best efforts fail (i.e. ozone treatment, kernel sorting). Mills association in Kenya was only operating at about 60%; but said if they had a de-contamination option, they would use it.
  • Informal markets of the supply chain must not be ignored.
  • How can the right behaviors be incentivized?

5. The challenges of managing aflatoxins globally involve aspects that are bigger than the scope of this coalition. Nevertheless, we must help find ways to engage global stakeholders to deliver the right support, to those who need it most

  • Support is needed from policymakers to ensure that predictive capabilities and modelling are adequately supported by effective policy and decision making at a country and global level.
  • Aflatoxin contamination and mitigation efforts impact all. In our efforts to help, everyone in the supply chain should be considered, from farmers (growers) to governments, empowering all.
  • All recommendations and decisions must be science led.
  • We should encourage grass-roots solutions. For example, to develop a Community of Practice with those who have dryers/storage/other innovations locally and engage with governments.
  • Another component for driving change could be improvement on current sourcing and pricing practices (are there any initiatives to standardize pricing?).

6. Education and communication must form part of any action taken to advance prediction and mitigation for aflatoxin contamination

  • There is a need for advocacy to show the benefits for a country to invest in data and evidence generation – for example, the possibility for a risk prediction model to better target control, risk analysis etc.
  • There is a need for a considered approach regarding communication of data and results (there are examples of where nations have suffered when high levels of aflatoxin have been made known, provided).
  • The current ‘risk-based’ approach to education and communications around aflatoxin risk management does not always reach the audience. This is for many reasons but there is a clear opportunity to improve education and communication efforts through a coordinated and consistent approach. “Telling the story” is important. A proactive and holistic approach is required.
  • Education is an important enabler – including the training of farmers, aggregators, and others.
  • Behavioral or social scientists could aid these efforts by helping create a deeper understanding of barriers to adopting new aflatoxin risk management strategies to help lay the path for effective mitigation in the food system.

7. There may be a role for Artificial intelligence (AI) and predictive analytics in the future

  • One use of the model may be to predict and understand mitigation steps (predictive and prescriptive analytics) if the data set is sufficiently large and robust.
  • It may be possible to look at the risk and the impact of various interventions, such as cleaning the corn or using different agricultural practices or storage mechanisms.
  • There should be caution and a note not to oversell what can be done with AI in the food safety space.

8. Measurement of food safety data and performance (through metrics) must focus on assuring lasting and sustainable change

  • Truly understanding the data that are available in food safety will require significant investment.
  • There must be a mutual, sustainable, and tangible Return on Investment (ROI), meaning that efforts must be focused to ensure the positive and lasting impact in food safety risk management that is being sought.

9. Options for expanding the aflatoxin model to include more data

  • The aflatoxin model has been tested and scientifically validated in several countries including India and Thailand. Additional, anonymized data is needed to help with refinement of the model and broaden aflatoxin mapping capability.
  • It is important not to shift the focus away from regions where it has been historically difficult to obtain clarity and data.
  • The following African countries were identified as possibilities for a pilot:
    • Nepal, Bangladesh, Malawi, Uganda, Ghana, Nigeria, Kenya
Food Safety Coalition Workshop: Actions and Next Steps

The coalition agreed to move forward in the following areas (below), while elevating the importance of food safety culture, regulation, and policy (items beyond the scope of this group) to the UN Food Systems Summit CEO consultations.

The coalition team has agreed to split into smaller work groups to develop actionable plans that can be driven at pace right away in the following areas. Updates on the group’s progress will be shared in the run up to the UN FSS in September and implemented on an ongoing basis:

  • Work stream 1: Sampling and data analysis best practice framework, education recommendations and pilot
  • Workstream 2: Risk assessment of what the data tells us (pilot and analysis; report and recommendations)
  • Work stream 3: ‘Test and learn’, creating case studies in select countries (pilot using model, underpinned by government engagement, and education – workstream 4)
  • Workstream 4: proactive social, education and communications pilot (link to workstream 3)

Members of the coalition will be notified on next steps over the coming weeks based on the agreed outcomes of the initial workshop.