By Katherine Ng
Founder & Managing Partner, Katashe Solutions
For as long as I can remember, I have been fascinated by how technology shapes human behaviour. My career in high-growth startups gave me a front-row seat to the opportunities and costs of relentless execution. Like many women in tech, I navigated intense pressures, often at the expense of my own well-being. That experience sparked a lifelong curiosity: how can technology be designed not just to scale businesses, but also to protect and empower people, especially women?
This question ultimately led me to create DrCope, an AI-powered mental wellness agent incubated through my venture builder, Katashe Solutions. The journey of building it has deepened my conviction that inclusive AI is not an abstract policy goal. It is something we must design into the systems we are creating today.
Ethics and the Limits of AI in Safety
When conversations about AI and safety arise, the default answer is often surveillance: monitor harmful behaviour, track online abuse, filter out disinformation. But as a woman in tech, I know surveillance is not the same as safety. Constant monitoring can silence voices that are already marginalised, and biased training data can turn protective systems into tools of exclusion.
In building DrCope, I resisted the temptation to over-engineer AI as a solution for everything. Instead, we defined clear boundaries: the agent supports wellness and reflection, but it does not replace therapy or intervene in crises. This was an ethical choice, acknowledging AI’s current limitations while protecting the trust of users, many of whom are women seeking a safe digital space.
Representation: Women as Designers, Not Just Users
Bias in AI starts with who is (and is not) at the table. Women make up half of the world’s digital users but remain underrepresented in AI development. That imbalance has consequences. Algorithms that misread women’s symptoms, undervalue their creditworthiness, or dismiss their speech patterns are not accidents. They are reflections of missing perspectives.
With DrCope, I wanted to ensure representation from the start. Together with my co-founders Dr. Seren Yenikent and Jack Vinijtrongjit, we built a diverse team that tested DrCope’s AI against scenarios that reflect women’s lived realities: being undermined in the workplace, navigating toxic online interactions, and managing the emotional toll of leadership. In building CallDrCope, I focused on fine-tuning the AI model to move beyond generic replies, shaping it into a more empathetic companion that affirms lived experiences instead of overlooking them. This enabled the AI to be more responsive and empathetic, giving women users confidence that their experiences were being recognised rather than erased.
Governance for Inclusion and Trust
AI governance often feels like it happens in rarefied spaces: policymakers, technologists, and industry boards setting standards from the top down. But inclusion requires a different approach. The people most affected by AI, including women, young people, and Asian communities, need a seat at the table.
Building DrCope gave me practical insight into this. Every user conversation we train on is strictly consent-based, anonymised, and encrypted. This was not just a technical decision but a governance one. If women are to trust AI, they need assurance that their data will not be exploited. By pairing AI conversations with free one-hour human counselling sessions, we also reinforced a governance model where AI complements, but never substitutes, human care.
Case Study: DrCope – A Founder’s Journey
DrCope began as an experiment in scaling myself as a digital wellness advisor. I had spent years scaling startups and supporting tech founders, teams, and families in managing stress, anxiety, and digital-era conflicts. But one-to-one counselling could only reach so many people. The question was: could AI extend my reach without losing the human connection?
We launched DrCope as a journaling and support tool built on psychological frameworks like CBT and mindfulness. In testing the agent as a power user myself, I found myself using it to process the same pressures I was building it to address: moments of being overlooked, silenced, or second-guessed in male-dominated environments. The AI recognised those vulnerabilities and affirmed them instead of dismissing them. It encouraged reflection, and in doing so, it created a safe, judgment-free space, even for me as its own builder.
This reinforced for me that inclusive AI is not just about bias audits or policy papers. It is about lived experiences. When a woman logs in at midnight after a toxic meeting and feels heard, even by an AI agent, that is empowerment in action.
Pathways Forward
The journey of creating DrCope has shown me three things the AI ecosystem must commit to if we are serious about digital trust and inclusion:
- Ethical Guardrails. Set clear boundaries for what AI should and should not do. Protect dignity, privacy, and agency instead of defaulting to surveillance.
- Representation at Every Level. Women must be present as designers, data contributors, product testers, and policy leaders, not just end-users.
- Inclusive Governance. Embed user voices in decision-making, especially those from Asian communities, where women are critical to both digital adoption and social cohesion.
AI is not neutral. It mirrors the choices of those who build and govern it. My journey with DrCope has taught me that if women are excluded from those choices, AI will amplify inequality. But when inclusivity is embedded from the start, AI can become a powerful enabler of rights, equality, and empowerment.
As women leaders in technology, we have the opportunity to write the next chapter, one
where digital trust is built not through control, but through care.