The 10 most innovative women in Artificial Intelligence

Updated on: March 3, 2026

Who is behind the results generated by AI?

If AI decision engines today automate critical flows, indicating who should receive credit, flagging potential fraud, or classifying risks that deserve attention, a fundamental question often goes unnoticed: who is behind the rules and models that underpin these automations?

Despite representing more than half of the world's population, women are still a minority in the scientific and technical production that underpins Artificial Intelligence. According to the UNESCO report (2025), women occupy approximately one-third of science and technology research positions worldwide, an indicator that remains below gender equality, even after decades of initiatives to promote female participation in these fields.

The report itself points out that, although women often perform academically on par with men in many higher education courses, their presence declines as their careers progress. This decline is especially noticeable in leadership positions and in areas such as engineering, technology, and information technology.

This pattern is not unique to academia. According to TechTarget (2025), data from 2024 indicates that women occupy about 27% of technology jobs globally, despite representing almost half of the total workforce. In the United States, this participation stood at around 35% at the end of 2023.

In Brazil, the asymmetry is even more pronounced. A study by the Softex Observatory, published by Softex (2025), shows that only 19.2% of information technology specialists are women, even though they already make up the majority of the population and a significant portion of the formal labor market.

The cumulative effect of these structural inequalities also appears in long-term recognition: women represent only a small fraction of Nobel Prize-winning scientists throughout history, which helps illustrate how barriers to access, permanence, and visibility extend across generations.

In recent years, training and inclusion initiatives have been attempting to reduce this gap. Large technology companies have begun investing in specific programs for women in AI, such as #ElasNaIA, from Microsoft, which offered training opportunities aimed at this audience. These movements are still insufficient, but they indicate a change in awareness in the sector.

Despite the barriers, many women have stood out and directly influenced the direction of AI, not only by creating technology, but also by expanding the way the market understands data, decision-making, and impact. These trajectories show, in practice, how technical and organizational choices shape systems that operate under uncertainty, risk, and real consequences.

One Brazilian example is Manoela Morais, who, alongside Chimka Munkhbayar and Helen Tsai, co-founded Agrolly, a platform that supports farmers in crop planning and climate risk mitigation. The application of AI in this case is not abstract: it acts directly on climatic and environmental variables that impact agricultural planning. In 2021, Manoela was recognized by the Women Leaders in AIprogram, which highlights women using AI in different sectors. 

 

Models extract patterns, not values

These figures are not just social statistics. They help explain why certain systems fail in predictable ways, especially when operating in complex, dynamic, and sensitive contexts.

A classic example came from Amazon in 2018. The company discontinued an automatic resume screening system when it realized that the model penalized female applicants for technical positions. The reason was not "malicious intent" on the part of the algorithm, but something more structural: the model was trained with patterns extracted from historical data that was predominantly male. Thus, the model only reinforced existing patterns, with mathematical efficiency and real consequences.

This type of situation raises an important distinction: models extract patterns, not values. When data carries historical asymmetries, models tend to reproduce them, unless someone notices, questions, and intervenes.

As AI decision engines become established as support for frequent , this dynamic takes on another scale. In areas such as credit granting, risk classification, or fraud detection, small statistical distortions are no longer marginal and consistently affect thousands or millions of people. The problem is not just "making mistakes," but making mistakes in the same direction, silently and difficult to contest.

Cases such as Apple Card have made this mechanism more visible. By generating significantly different credit limits for men and women with similar or even superior financial profiles, the system highlighted how automated decisions can incorporate historical inequalities without this being explicitly coded. As pointed out at the time by reports in The Guardian (2019), the difficulty in explaining the model's criteria exposed the so-called "black box" of algorithmic decision-making, where not even its users or creators can clearly identify how certain results are produced.

In regulated contexts, this scenario is particularly sensitive. Unfair or unrepresentative criteria, when incorporated into models that operate continuously, tend to crystallize as stable operating rules. Without systematic human review, adequate metrics, and diversity in the teams that design, validate, and govern these systems, AI not only reflects the past, it automates it in the present, at scale.

It is precisely at this point that the presence of women in the development, application, and governance of AI ceases to be merely a matter of inclusion and becomes technically relevant. Expanding the pool of people who ask questions, define data, and question results is one of the most effective ways to reduce persistent biases in automated decisions. 

The following trajectories help to make this role visible.

 

To understand how these systems evolve beyond static models, it is worth reading the article: Adaptive AI: The Evolution of Traditional Artificial Intelligence

 

Meet the 10 most innovative women in the Artificial Intelligence industry

 

10) Rana el Kaliouby

An Egyptian-American computer scientist, Rana holds a PhD from Cambridge University and a postdoctoral degree from MIT. She co-founded Affectiva, a pioneer in Emotion AI, and is currently Vice President of AI at Smart Eye. Her career is dedicated to humanizing technology by integrating emotional intelligence into data processing.

Rana has developed systems that enable models to classify patterns of expressions associated with complex human emotions through facial expressions and tone of voice. In practice, this has revolutionized everything from vehicle safety (detecting driver fatigue) to mental health monitoring and market research, proving that "data" can also be feelings.

In the current scenario, Rana represents a break from the "cold machine" paradigm. She argues that for AI to be truly intelligent and useful in critical decisions, the system needs to be calibrated to process human context and emotional state. In the world of critical decisions, Rana's work shows that ignoring human emotional context is a technical error. Her work redefines the boundary between signal processing and empirical understanding, connecting human sensitivity to the robustness of algorithms.

TED Talk recommendation: This app knows how you feel — from the look on your face References: Dr. Rana el Kaliouby – Personal website Harvard Business School – Professional Profile World Economic Forum – Professional Profile SpringAct – Inspiring Women in Tech

 

9) Shivon Zilis

With a degree in Economics and Philosophy from Yale, Shivon is one of the strategic minds behind Neuralink, where she serves as director of operations and special projects. With stints at Tesla and OpenAI, she has established herself as an elite investor and executive in the Deep Tech ecosystem.

Shivon focused her contribution on how Artificial Intelligence can be symbiotically integrated with human potential, especially through the brain-machine interface. She helped direct investments and architect the growth of companies that today define the state of the art in neural network models, accelerating the transition from theoretical concepts to real industrial and biomedical applications.

Today, she represents the necessary bridge between business strategy and technological feasibility. By questioning how superintelligence will be governed and integrated into biology, Shivon brings the perspective of security and scale to the debate. Her presence is relevant in ensuring that innovation is not only powerful, but also safe and integrated into human biology in an ethical manner.

 

8) Anna Patterson

With a PhD in Computer Science from Illinois and a robust background as Vice President of Engineering at Google, Anna Patterson founded Gradient Ventures to drive the AI startup ecosystem. She was one of the key architects of Google's search engine, leading teams that designed data indexing systems on a planetary scale.

When she transitioned to investing, Anna began shaping the market by selecting and mentoring companies that solve complex problems in sectors such as healthcare and logistics. Her work enabled AI solutions to move from the drawing board to optimizing real workflows, ensuring that the technology reached the end user efficiently and with technical robustness.

Anna is proof that modern AI depends on infrastructure fundamentals. She broadens the debate on how to transform academic research into tools that work under pressure in the global market. Her vision reinforces that any high-impact automated decision is only possible if there is a flawless data architecture behind it.

References: VC Corner: Anna Patterson, Gradient Ventures – Startup Grind Introducing Gradient Ventures – Google Blog

 

7) Kamelia Aryafar

Kamelia Aryafar holds a PhD in Computer Science and has built a solid career as CTO at Overstock and AI leader at Google Cloud. An expert in Machine Learning, she focuses her work on the intersection between complex mathematical models and immediate business results, especially in retail and logistics.

She transformed the way large platforms manage inventory and personalize the user experience, using predictive models to generate recommendations that optimize supply chains. Her work has enabled e-commerce to operate with less waste and greater precision, solving logistical problems that were previously dealt with in a purely reactive manner.

In the context of automated decisions, Kamelia is a fierce advocate for explainability. She questions the "black box" model, arguing that algorithms in business environments need to be transparent and generate real value for the customer. Her leadership redefines the role of AI in operations, combining statistical rigor with real-world efficiency.

References: Kamelia Aryafar – Professional profile and experience Drexel Magazine – Academic background and e-commerce Drexel CCI – Advisory council and roles in technology

 

6) Poornima Ramaswamy

A high-level executive with strategic stints at Cognizant and Qlik, Poornima Ramaswamy has established herself as an authority on Digital Transformation and Data Strategy. She works on the front lines helping global companies structure their data governance so they can finally use AI responsibly.

Poornima focuses on the "end of the line": how companies use insights to make decisions. She has implemented governance frameworks that have enabled global companies to migrate from intuition-based decisions to a rigorously data-driven culture. By focusing on the integrity and quality of information, she has helped reduce operational and financial risks by ensuring that AI-generated insights are reliable and auditable.

Her presence is vital to the debate on data governance. She redefines the role of the technology leader as the guardian of trust in automated decision-making. In a world where data can be biased, Poornima's work ensures that companies have the necessary processes in place to question and validate what the machine suggests.

References: Qlik press release – Poornima Ramaswamy appointed EVP, Global Solutions and Partners Qlik Blog – Poornima Ramaswamy insights and executive profile CRN Channel Chiefs 2022 – Influential channel leaders

 

5) Fei-Fei Li

A professor at Stanford University and co-director of HAI (Human-Centered AI), Fei-Fei Li is recognized worldwide for creating ImageNet. This massive database was the fuel needed for the explosion of Deep Learning and modern computer vision, allowing models to be trained to identify patterns in the visual world.

Her contribution has radically changed sectors such as healthcare, enabling more accurate diagnostic imaging, and transportation, through the development of autonomous cars. During her time at Google Cloud, she brought this cutting-edge science closer to business solutions, proving that AI needs diverse data to be truly effective.

Fei-Fei is the leading voice of "Human-Centered AI." She questions technological development that is detached from ethics, arguing that technology should be designed to collaborate with people, not just replace them. Her vision is what underpins today's debates on diversity and inclusion in the training of large models.

TED Talk recommendation:  How we’re teaching computers to understand pictures The jobs we’ll lose to machines — and the ones we won’t (participation) References: Stanford University – Academic and professional profile Stanford Vision and Learning Lab – Research and laboratories

 

4) Daniela Rus

Director of MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL), Daniela Rus is one of the world's leading authorities on robotics. With a rigorous background in mathematics, she leads research exploring how artificial intelligence can be incorporated into machines capable of acting and executing autonomous commands physically in the real world.

She has developed robots capable of autonomously reconfiguring themselves and AI systems that enable fluid interaction between machines and humans in industrial environments. Her innovations help solve productivity bottlenecks in critical areas such as logistics and manufacturing, where the accuracy of the outputs generated by the model translates into direct physical action.

Daniela redefines intelligence as something "embodied." Her work shows that AI decisions do not end with software; they manifest themselves in movements in the physical world that require absolute safety. She expands the industry's technical repertoire by proving that autonomy requires a level of mathematical rigor that goes beyond text or image processing.

TED Talk recommendation: Meet the robots for humanity References: Daniela Rus – Personal website MIT – Academic profile of Daniela Rus IEEE – Edison Medal 2025 – Scientific recognition

 

3) Joy Buolamwini

A researcher at MIT Media Lab and founder of the Algorithmic Justice League, Joy Buolamwini uses her technical background to audit the integrity of AI systems. Her research, "Gender Shades," became a global milestone by proving that facial recognition systems from large companies had critical flaws when identifying women and black people.

Her work forced a complete overhaul of how technology companies train and validate their models, leading giants such as Amazon and IBM to suspend biased surveillance technologies. Joy transformed algorithmic auditing into a tool for social justice, ensuring that technical flaws do not become tools of exclusion.

Joy represents the pillar of transparency and ethics. She questions the paradigm of mathematical neutrality, showing that if the data contains historical asymmetries, the model will replicate these statistical patterns. Her presence is what ensures that the AI of the future will be built with a critical eye that prioritizes equity in automated decisions.

TED Talk recommendation:  How I’m fighting bias in algorithms References: Algorithmic Justice League – Organization and mission MIT Media Lab – Academic Profile Gender Shades – Study on bias in facial recognition TIME – 100 Most Influential People in AI (2023) – International recognition

 

2) Mariah Scott

Mariah Scott is an executive specializing in managing emerging technologies in highly regulated markets. After leading Skyward (Verizon), she consolidated her work in applying AI to the management of critical infrastructure, such as electrical grids and large engineering systems.

She enabled the use of fleets of drones and smart sensors to perform autonomous inspections, transforming predictive maintenance into an operational science. By developing systems that indicate the ideal moment for intervention before a failure occurs, her work directly impacts public safety and resource savings in high-risk sectors.

In the debate on AI, Mariah represents the pragmatism of operations. Her work shows how decision models applied to infrastructure save lives and optimize national logistics. She redefines AI as a tool for resilience, capable of keeping essential services running in extremely complex environments.

References: GSMA – Mariah Scott (President, Skyward, a Verizon company) Telecom TV – Verizon launches Robotics Business Technology

 

1) Lisa Su

Lisa Su is the engineer and CEO who led AMD's transformation into one of the world's most valuable companies. With a PhD in Electrical Engineering from MIT, she used her deep technical knowledge to redesign the architecture of processors and GPUs, placing them at the center of the global AI infrastructure.

Without the high-performance chips Lisa designs, training models such as ChatGPT and complex fraud detection systems would not be possible. She has democratized access to heavy processing, allowing AI to move out of laboratories and run on supercomputers and data centers around the world with energy efficiency.

Lisa Su is the material foundation of the technological revolution. She reminds us that every software decision ultimately depends on an atom of silicon. Her leadership redefines technological sovereignty and ensures that the infrastructure needed for the future of AI is robust, powerful, and capable of supporting tomorrow's critical decisions.

References: AMD – Executive Leadership and Corporate Vision Massachusetts Institute of Technology (MIT) – Academic background and career in engineering Semiconductor Industry Association (SIA) – Institutional leadership in the semiconductor industry

 

---

About 4kst 

4kst is a Brazilian DeepTech company born at PUCPR, a pioneer in the development of Adaptive AI. Through proprietary Data Stream Learning technology, we create predictive models that learn and update in real time. Unlike traditional Machine Learning, our solution eliminates performance degradation and reduces maintenance costs. Two-time winner of Febraban Tech and recognized by Finep, 4kst combines cutting-edge science and high performance to keep your company ahead in dynamic markets.