In today’s digital age, the exponential growth of data from diverse sources has paved the way for remarkable advancements in artificial intelligence, machine learning, and predictive analytics. Amidst this rapid evolution, ethically-aligned AI has garnered significant attention, becoming a crucial concept that encompasses the development and deployment of AI systems aligned with ethical principles and values.

One of the major concerns surrounding AI is its potential to perpetuate biases or discrimination present in the data it is trained on. For instance, facial recognition systems have been found to exhibit racial and gender biases due to biased datasets. Moreover, the collection and use of personal data raise privacy concerns, increasing the risk of misuse. Former Google Design Ethicist Tristan Harris has emphasized the importance of coordinating efforts and implementing safety measures in AI development to avert potential tragedies. The complexity of balancing specific objectives with broader societal goals, such as fairness and safety, poses challenges for institutions in handling ethical considerations.

As AI technologies become more pervasive across industries, the need for ethical considerations in data handling and usage has grown exponentially. This has led to the emergence of the ethical AI startup ecosystem, with notable players like the Ethical AI Database (EAIDB), the only publicly available, vetted database of AI startups providing ethical services, offering market maps and reports published semiannually, and continually updating its database.

Abhinav Raghunathan, the founder of EAIDB, shared that the birth of this initiative was motivated by the need to draw attention to an underserved startup space that aimed to make a real impact by mitigating the downsides of AI/ML (artificial intelligence/machine learning). He said that the motivations behind the market research for ethical AI are multidimensional. For example, AI investors are increasingly seeking to assess AI risk as a fundamental aspect of their comprehensive profiling of AI companies. This trend is supported by the growing demand for ethical AI practices identified in the IBM Institute for Business Value’s report. Large enterprises are recognizing the need to operationalize, quantify, and manage AI risk. To achieve these ends, internal risk and compliance teams are diligently identifying suitable toolsets for effective risk management. Abhinav also pointed out that beyond practical considerations, a profound philosophical perspective underlines AI’s development. AI systems should be designed to work for everyone, not just a specific segment of the population. Ensuring fairness and transparency in black-box algorithms and opaque AI systems is of the utmost importance to uphold ethical standards in AI applications.

One of the major concerns surrounding AI is its potential to perpetuate biases or discrimination present in the data it is trained on.

When it comes to selecting the right candidates for their database, EAIDB is dedicated to eliminating companies that merely profit from the buzzword “responsible.” They meticulously source, categorize, and research ethical AI enablement solutions. The stringent verification process ensures that companies featured in the database are actively involved in promoting responsible AI practices, enhancing transparency, and embracing accountability within the rapidly evolving AI industry.

For EAIDB, an “ethical AI company” is one that either provides tools to make existing AI systems ethical or develops products that address bias, unfairness, or unethical elements in society. Their comprehensive approach has identified eight distinct categories of responsible AI startups, including Data for AI, MLOps and ModelOps, AI GRC, Model and Platform Builders, AI Security, Alternative ML, Consulting, and Open Source.

For EAIDB, an “ethical AI company” is one that either provides tools to make existing AI systems ethical or develops products that address bias, unfairness, or unethical elements in society.

To maintain the database’s consistency, the verification procedure for the Ethical AI Database (EAIDB) entails several prerequisites, with a primary focus on early-stage companies founded post-2015. This strategic choice aligns with the understanding that direct innovation in the AI industry tends to happen at this stage, where startups are at the forefront of cutting-edge developments. Consequently, companies beyond Series C funding are excluded.

Inclusion in the EAIDB hinges on a company’s core business aligning with the concept of “responsible-enabling,” implying a dedication to contributing to responsible AI practices. The verification process is thorough and seeks evidence of responsible activities, such as the presence of active research teams working on explainability and fairness. Additionally, EAIDB evaluates products or services that align with the eight responsible AI categories, along with the level of corporate social media engagement on responsible AI topics. A company showing sufficient evidence becomes “indirectly verified,” while to achieve “direct verification,” the EAIDB team engages in meetings with founders to discuss technical details and, at times, witness demos.

The founding of EAIDB was marked by collaboration with other significant players in the ethical AI landscape, such as EAIGG (Ethical AI Governance Group), which provides best practices for open-sourcing ethical AI governance, and BGV (a cross-border VC platform). Since its inception, EAIDB has closely monitored and tracked over 260 companies, engaged with foreign government institutions, and generated over 3,000 reads on their reports. Their dedication has been instrumental in helping companies featured in the database access VC funding and clients. Startup executives are increasingly recognizing that building ethical AI solutions not only aligns with a moral imperative but also offers strategic advantages. Investors are also acknowledging the value of ethical AI companies, appreciating their potential to gain public trust and mitigate potential legal and reputational risks.

With data becoming increasingly influential, those with access to vast amounts of data and the ability to analyze it hold the power to make informed decisions, predict future outcomes, and shape strategies. As AI technology progresses, it is crucial to strike a delicate balance, leveraging the power of data for positive advancements while ensuring responsible and ethical use. EAIDB acknowledges that their journey comes with challenges proportional to the mission’s importance. The team is committed to constant learning, adaptation, and collaboration for effective course correction. EAIDB is not immune to pushback concerning the definition of “responsible-enabling.” Moreover, they face the difficulty of managing limited human resources to scale rapidly on new technologies, thereby enabling a better profiling of the ethical AI startup landscape. Despite being a small team, they have directly verified around 45% of the full database and remain committed to increasing this number.