Who Makes the Rules for AI? Careers, Policy, and Accountability in the Age of AI

Admin ILTN
Jul 20
7 min read

AI is developing faster than the rules meant to govern it. While governments, companies, researchers, and civil society debate what responsible AI should look like, the technology is already being deployed across sectors and becoming part of everyday life.

That tension was at the centre of a recent Indian LegalTech Network session on careers, policy, and accountability in the age of AI.

The conversation brought together Jameela Sahiba and Vaishali Gopal for a candid discussion on what it means to work in public policy, how AI governance is taking shape in India, and why more young professionals, particularly women, need to participate in these conversations.

The session also offered a realistic look at policy careers. There is rarely a straight path, immediate impact, or fixed rulebook. Much of the work involves learning to operate in ambiguity, building relationships, bringing different stakeholders together, and staying committed even when the policy agenda changes.

There is no single route into public policy

One of the clearest takeaways from the session was that a career in public policy does not always begin with a policy degree or a perfectly planned transition.

Law students are generally presented with a familiar set of options: litigation, law firms, in-house roles, judicial services, or the civil services. Public policy is often missing from that list, even though legal training can provide many of the skills the field requires.

A background in litigation, legal research, competitive examinations, or legislative work may appear disconnected. In practice, each experience can build a useful part of the policy toolkit: research, writing, close reading, institutional understanding, and the ability to construct an argument from incomplete information.

The larger lesson was that careers do not have to make perfect sense while they are unfolding. Sometimes an unsuccessful attempt, an unexpected opportunity, or a role outside the traditional legal path becomes the turning point.

Moving from law to policy requires a different mindset

Legal work often begins with a defined problem. A lawyer is asked to research the applicable law, identify precedents, and develop the strongest possible argument.

Policy work is less predictable.

A policy professional may be asked to examine a national issue, understand a problem affecting a particular constituency, prepare a legislative briefing, speak to industry representatives, or identify how a proposed law will affect different groups.

There may not be a clear right or wrong answer. Different stakeholders experience the same policy differently, and an intervention that works for one group may create difficulties for another.

The session described this as learning to work with a clean slate. The shift is from asking only, “What does the law say?” to asking, “What is the problem, who is affected, what evidence do we have, and what could work in practice?”

That requires comfort with grey areas. Policy is often experimental, and clarity develops gradually through research, consultation, and revision.

Research alone does not shape policy

Another important insight was that producing good research is only one part of policy work.

A report does not influence the public conversation merely because it has been published. The findings must reach the people who can use them: legislators, government departments, industry leaders, civil society organisations, researchers, and affected communities.

Policy professionals therefore act as translators and bridges. They explain technical questions to legislators, institutional limitations to industry, public harms to companies, and political realities to researchers.

This may involve preparing briefings, meeting stakeholders, organising closed-door discussions, presenting research, supporting public interventions, or helping different groups understand one another’s concerns.

The discussion made one point especially clear: stakeholder engagement is not separate from policy work. It is a central part of how policy is made.

Networks matter, but credibility matters more

Networking emerged as a recurring theme throughout the session.

In public policy, relationships can lead to opportunities, collaborations, and access to conversations that may not be available through formal applications. They can help research move beyond the organisation that produced it and reach the institutions capable of acting on it.

But meaningful networking is not about collecting the largest possible number of contacts. It is about building credibility over time.

That credibility comes from understanding what someone is working on, approaching them with something useful to offer, delivering what was promised, and remaining in touch without making every interaction transactional.

Not every relationship will produce an immediate opportunity. Someone who appears unrelated to a current project may become relevant several years later. There is no perfect formula for deciding which connection will matter.

The practical advice was simple: attend events, speak to people across the ecosystem, maintain an active professional presence, explore possible areas of collaboration, and allow trust to develop through the quality of the work.

Access may begin with an introduction, but reliability is what makes the relationship last.

Women should not assume that policy spaces are closed to them

The conversation also addressed the experience of women entering political and policy spaces.

These spaces are still largely male-dominated, and the idea of working closely with political offices or senior public figures can create apprehension, both for young professionals and their families.

But the session challenged the assumption that these spaces are necessarily inaccessible. Women can build trusted, professional relationships across institutions and political lines when they enter with confidence in their skills and demonstrate value through their work.

Confidence does not mean knowing everything. It means being clear about what one can contribute, proposing useful ideas, and following through consistently.

The wider concern remains important: if women and other underrepresented groups are absent from policy conversations, their experiences may also be absent from the rules that eventually emerge. Participation is therefore not only a career question. It is also a governance question.

AI safety is becoming a central policy concern

The discussion then moved from policy careers to the substance of AI governance.

One of the most urgent issues is safety, particularly for children, young adults, and groups that may be disproportionately affected by technology.

Deepfakes, misinformation, child sexual abuse material, harmful content, and addictive online behaviour are no longer hypothetical concerns. As AI becomes integrated into social media, education, finance, healthcare, and public services, questions of safety and accountability will become harder to separate from questions of innovation.

This does not require an anti-technology position. AI can improve services, expand access, and create new opportunities. But its benefits cannot be considered without also examining who carries the risk when a system causes harm.

The more useful questions are: Safe for whom? Who is missing from the conversation? What happens when an automated system fails? And what remedy is available to the person affected?

Young people should not be treated only as users who need protection. They should also have opportunities to participate in defining what safe and beneficial technology looks like.

Responsible AI needs more than isolated conversations

The session also examined how coalitions and professional communities can contribute to responsible AI.

Different parts of the ecosystem often discuss the same issue separately. Companies speak to other companies. Start-ups speak to start-ups. Researchers, civil society organisations, and public institutions may each remain within their own circles.

The problem is not always the absence of discussion. It is the absence of meaningful exchange between these groups.

Building an effective coalition begins with identifying that gap. The next step is stakeholder mapping: who is affected by the problem, who has relevant expertise, and who can influence the outcome?

Only then should organisers create a blueprint covering the coalition’s purpose, principles, membership, working structure, and intended outputs.

Even after launch, the original plan will need to change. New members will identify missed assumptions, raise practical questions, and sometimes send the organisers back to the drawing board.

The takeaway was that successful communities are not built from a perfect first plan. They are built through clear purpose, careful groundwork, and a willingness to learn as the work develops.

Policy impact rarely arrives quickly

One of the most honest parts of the session concerned frustration.

Policy work does not always produce visible or immediate results. A professional may spend months researching one issue only to find that the government’s focus has moved elsewhere.

AI policy makes this especially clear. The conversation can shift from broad principles of responsible AI to sector-specific risks, AI safety institutions, domestic capacity, or technological sovereignty. The goalpost keeps moving because the technology and the political priorities surrounding it are also moving.

This can create the feeling that earlier work has been wasted. But policy impact is not limited to a recommendation becoming law.

Research can introduce new language into a debate, prepare a legislator for a future intervention, bring an excluded stakeholder into the room, or help an institution understand an issue before it becomes urgent.

The result may be delayed, indirect, or difficult to claim. Resilience in policy means recognising that influencing the process can be valuable even when the final outcome is not immediately visible.

What young professionals can do

For students and young professionals interested in technology policy, the session offered several practical starting points.

Internships with technology policy think tanks and civil society organisations can provide visibility into how these institutions actually work. Even short experiences can help someone understand the research process, stakeholder engagement, capacity building, and the different kinds of outputs policy organisations produce.

Writing publicly can help develop expertise and make a person’s interests visible. Attending events and reaching out thoughtfully to professionals can create a better understanding of the ecosystem.

Most importantly, young professionals should not wait until they feel like complete experts before participating.

Emerging fields often favour people who are curious, prepared, and willing to learn while the space itself is still developing.

Final takeaway

The biggest takeaway from the discussion was not that AI policy belongs only to governments, technologists, or senior experts.

It belongs to everyone who will be affected by the systems being built—and that means the conversation needs more lawyers, researchers, students, women, civil society organisations, industry professionals, and public voices.

The rules for AI are being written while the technology is already in use. There is no finished map, and there may never be a single moment when the work feels complete.

But that uncertainty is also an opportunity.

The people who enter these conversations now, build credibility, ask difficult questions, and bring others into the room will help determine what responsible AI eventually means in practice.

The rulebook is still being written. The question is who chooses to help write it.