Is Your Hiring Network a Hidden Filter?
In fast-moving fields, companies hire quickly through familiar networks. The speed that feels efficient can quietly narrow your talent pool.
AI should, in theory, be a fair field. Few people have a decade of experience with technology that is a year old. Yet women working in AI told The Guardian that hiring has leaned on personal networks and loose signals instead (Pardes, 2026). The head of a non-profit advancing women in tech described companies hiring at pace through “the same networks, the same referrals, the same filters” (Pardes, 2026).
Is your network a filter?
The numbers show the result. Women made up about a quarter of new AI hires in the past year, compared with 50% of new hires in non-AI roles. In executive AI roles the figure was 13% (Pardes, 2026).
Retention is the second half of the problem. The article reports that women leave the tech industry at higher rates than men, citing non-inclusive cultures and barriers to advancement (Pardes, 2026). One AI engineering lead described returning from leave to find tools and models had changed completely, and said women may lack the “infrastructure that makes keeping up possible” instead of the ability or interest (Pardes, 2026).
SMEs face the same pressure on a smaller scale. When a skill is scarce, the fastest route is a referral. When the workload is intense, the people with the most support at home can keep up and others can fall behind.
Implications for SMEs
Widen the funnel. Advertise roles beyond personal networks and state the skills required in plain terms.
Use structured hiring. Ask every candidate the same questions and score against agreed criteria.
Plan for re-entry. After any extended absence, offer a structured catch-up plan. Fast-changing tools make this a productivity measure as much as a fairness measure.
Track exits. Review who leaves and why. Patterns often show up before a formal complaint does.
Narrow hiring and weak retention both reduce the range of thinking available to the business, which is a talent risk with a direct cost.
Limits. The article is journalism and includes the views of advocates. It also describes a US policy context in which diversity programs have been scaled back and challenged (Pardes, 2026). Australian SMEs should check current local legal requirements before designing any program.
References
Pardes, A. (2026, 8 October). Women hold just a fraction of new AI jobs but are overrepresented in roles with high risk of AI disruption. The Guardian.
