We systematically implement and apply this statement of values in our projects. Our main values :
Algora Lab is motivated by democratic values. We are a space for deliberation on the ethical, societal and political impacts and challenges of learning algorithms and digital innovation, bringing together stakeholders, experts and non-experts, representatives of civil society organizations, the private sector and public policy makers. We uphold inclusiveness, open-mindedness and tolerance during our consultation and deliberation processes.
Algora Lab will remain transparent in its actions, reporting to other partner institutions and making our work available on our website. We will disclose all sources of funding and financial support. In addition, we will limit corporate funding to 15% of Algora Lab budget.
A supervisory committee will ensure that Algora Lab’s research projects remain transparent and ethical. It will be composed of international experts and well-known personalities in the fields of ethics and artificial intelligence.
Algora lab’s actions are based on the principles of capacity building and collaboration. We are committed to empowering communities through consultation and deliberation. These two methods are at the heart of our action, but are not sufficient on their own to guarantee an effective empowerment. Therefore, we also offer training to educational and research institutions, public and private organizations on the ethical issues of AI and data use, and more specifically on the Responsible AI Montreal Declaration and its application to the design and deployment of AI systems.
Algora Lab complies in all respects with the Montreal Declaration for a responsible use of Artificial Intelligence. In 2017, the Université de Montréal, with the support of FRQ (Fonds de la recherche du Québec), CIFAR, Mila and IVADO launched a participatory process leading to the elaboration of this document.More than 20 public events and deliberative workshops were organized and over 500 people were involved in the face-to-face and online deliberative workshops.
The scientific leaders of the Montreal Declaration with several research assistants who worked on it are part of the core team of Algora Lab.
4. Equity, Diversity, Inclusion & Gender Equality
We are committed to including diverse sets of participants, including ambassadors, employees. We guarantee a fair remuneration to all employees regardless of their country of origin or gender. This project should benefit the communities that will be involved, and all their members. Special attention and targeted strategies will be implemented to reach traditionally marginalized voices and ensure their inclusive and meaningful participation in the forum.
1. AI diversity crisis spans from academia to the industry. In academia, women account for only 18% of authors at leading AI conferences are women, and more than 80% of AI professors are men. The situation in the industry is equally worrying. Despite lack of data, other identity markers shape the AI sector (dimensions of race, LGBTQ people, people with disabilities, economically marginalized groups, youth or elderly, the “Global South”). The AI diversity crisis is essentially about power (where AI knowledge is basically concentrated in the hands of a few players).
2. AI international digital divide. According the 2019 Government AI Readiness Index, which provides an assessment of the level of preparedness of governments to benefit from the potential of AI, high-income countries with robust AI strategies and investments dominate the Index’s top 10 positions. “Overall, the Index indicated that governments in the Global North are in a better position to reap the rewards of AI than their Southern counterparts. If this imbalance of AI readiness persists, Southern countries will be at risk of becoming AI testing grounds where systems are misused”.
↳ “The disparities in AI readiness with the Global South could increase the risk of certain countries becoming testing grounds. Furthermore, AI could magnify existing inequalities if the unique context of each region is not considered”.
↳ “For instance, the growing use of facial recognition in Zimbabwe is happening without appropriate data protection policies, which is raising important ethical questions”.
3. AI bias and discrimination. In addition , “AI systems can reflect societal biases and assumptions held by their designers or inherent in the datasets on which their core algorithms are trained. Using biased systems to automate decision-making processes could amplify the impact of these biases by systematically producing results that disadvantage particular individuals and groups, especially those who are marginalized”: advertising targeting algorithms are more likely to offer women lower-paying job offers, AI working on language can reproduce sexist stereotypes; facial recognition technologies are more reliable with white men than with women and racialized people… Ultimately, people with few economic resources are those who end up paying the price for adopting decision systems. The possible consequences do not only affect individuals; loss of opportunities, economic losses and social stigmatization are among the risks faced by society as a community.
Several principles and recommendations of the Montreal Declaration for a Responsible Development of AI (which was born from an inclusive deliberation process) relates to equity, diversity and inclusion:
- Principles 5 – Democratic participation: AIS must meet intelligibility, justifiability, and accessibility criteria, and must be subjected to democratic scrutiny, debate, and control
- Principles 6 – Equity: The development and use of AIS must contribute to the creation of a just and equitable society.
- Principles 7 – Diversity inclusion:The development and use of AIS must be compatible with maintaining social and cultural diversity and must not restrict the scope of lifestyle choices or personal experiences.
“There is a collective responsibility for the social impacts of AI”
As the use of AI spreads, it is crucial that diversity be taken into account in the process of creating and determining the ethical principles that will guide its development and deployment. There is an urgent need to inform and consult citizens from diverse backgrounds to minimize potential harms, ensure AI technologies are used for the common good, and develop informed and participative policies.
Diversity of experience and perspectives is essential to the formulation of informed opinions and ethical choices. Algora rationale results from the need to “create inclusive means of consultations that involve citizens in all their diversity, at different steps in the oversight process for AI responsible development”11. For Algora, diversity includes obvious aspects such as origins, gender, age and socio-economic conditions, but also considers other essential elements such as culture, traditions and spirituality.
At Algora, we are aware that less-privileged and traditionally marginalized groups may be less apt to participate in deliberative events or feel less safe to express their views. Along with our partners, the Algora team takes concerted steps to mainstream diversity and inclusion in all aspects of their projects. This includes developing links with local communities in order to include under-represented groups in our activities and eventually, in the process of defining the ethical development and deployment of AI.
Do you want to get involved or know more about our commitment? Contact: Pauline Noiseau or Jihane Lamouri : Jihane.email@example.com
5. Confidentiality and Privacy
Algora Lab is committed to protect research participants’ privacy. Our events are public, but we protect participants’ identities – if they request it. We will not collect unnecessary personal information. The information we will collect, online or at the physical events, will not be shared with third-parties, unless beforehand explicitly stated.
6. Sustainable Development
Algora Lab is involved in international projects. As a general rule, we aim to reduce these travels to the necessary minimum. We also aim to proactively compensate the carbon footprint of our project.
Algora commitment to the equitable and inclusive development and deployment of AI
At Algora Lab, our research focuses on the ethical aspects and societal impacts of the use of AI. We therefore cannot disregard the environmental impacts of our research operations. A part of our expertise is to organize workshops and events internationally. Our activities have a significant environmental footprint, since they imply traveling, electronic equipment and organizing events.
We stand by the many scientific reports on the importance of mitigating climate change and avoiding the pollution generated by human activities. Algora Lab believes in the necessity to protect the environment and commits to limit and offset the greenhouse gas emissions as well as the waste generated by our activities.
Limiting our environmental impact
Most of our impacts are caused by organizing our events and traveling.
In our engagement to limit our environmental footprint, we commit to the following:
- Travel: For short distance mobility, we will promote using public transportation when possible.
- We will minimize traveling long distances. When necessary, we will choose the least carbon-emitting means of transportation. While keeping in mind logistical considerations, we will prioritize, in order: train, bus, car, boat and plane.
- Waste: During our events, as well as in our day-to-day activities, we will minimize as much as possible the amount of produced waste. We will design our workshops to reduce the use of disposable material, to reuse the material, and to recycle any waste that is nonetheless generated.
- Food: If food is an aspect of the events we organize, we will take the necessary measures to avoid wasting food, and ensure the remains are used for compost. Considering the high water consumption and greenhouse gas emissions due to the animal-based food industry, as well as the hardships it implies for sentient beings, we will prioritize plant-based products, preferably organic, when providing catering.
- In general: In any other context, we will always bear in mind the environmental impact of our activities and take measures to actively reduce our footprint.
Carbon offsetting: why and how?
Although we will make significant efforts to diminish our environmental footprint, the very nature of our activities and the necessary traveling will make it impossible to reduce it to zero. We therefore commit to quantify and offset the remaining impacts, especially with respect to greenhouse gas emissions.
We will offset our greenhouse gas emissions by financially contributing to projects which will either capture these gases from the atmosphere, or reduce emissions from another source. We acknowledge that carbon offsetting is not a perfect solution. However, we consider it to be the best option available when all reduction measures have already been applied.
We have selected two organizations to offset our emissions with equal contributions.
- Carbone Boréal is a Québec-based project led by researchers at Université du Québec à Chicoutimi (UQAC). The contributions are used to plant trees in the boreal forest, allowing to reliably and verifiably absorb carbon in the atmosphere, while constituting research plots that will be used by the University researchers to investigate carbon sequestration related topics. This platform was chosen for the double utility of the contribution (carbon absorption and scientific research) and the transparency of the process. Furthermore, we want to support a local initiative in the province of Québec, where Algora Labs is based.
- The United Nations Carbon Offset Platform allows to contribute to the development of UNFCCC certified projects in developing countries. The primary objectives of these projects are to reduce, avoid or remove greenhouse gas emissions from the atmosphere. Parallely, they also aim to improve the ecological, social and economic conditions in the region they are developed. This platform was chosen for the international outreach, the credibility of the UN program, as well as the positive impact of the contributions.
Both organizations set a price per ton of CO2 that the contributor wants to offset. The quantification of our greenhouse gas emissions will be made through the GHG calculator of Carbone Boréal, which is scientifically sound and allows to evaluate individually a wide range of elements, from different traveling modes to food, giving it the necessary flexibility to fit the nature of our events.
We also abide by the concept of Climate Positive actions as defined by the UQAC Chair of Eco-Advising, and commit to double our contributions for both carbon offset programs.
Algora Lab is an academic institution. We defend our intellectual and scientific independence. We are not a political or corporate organization, and we will remain so.
Ressources and date of revision :
We respect the Montreal Declaration for a Responsible Development of Artificial Intelligence: https://www.declarationmontreal-iaresponsable.com/. This statement of values will be updated after one year in December 2020.