Copy article

Women See 700% Reach Boost Switching LinkedIn Gender to Male

ended 15. December 2025

Women changing LinkedIn profiles to male reported impression increases up to 818%, according to multiple sources. Marketing leader Lucy Ferguson changed her name to "Luke" for 24 hours and saw impressions rise 818%. Writer Jessica Doyle Mekkes reported 700% jump changing only gender marker. Mental health professional Megan Cornish saw 400% page view increase after changing gender and rewriting posts in "more assertive, male-coded language."

Entrepreneur Cindy Gallop's post reached 801 people while a male colleague posting identical content reached 10,408 people, despite Gallop and Jane Evans having 154,000 combined followers versus the men's 9,400.

LinkedIn VP of engineering Tim Jurka announced in August 2025 the platform had implemented LLMs to surface content. Sakshi Jain, LinkedIn's head of responsible AI, stated "Our algorithms do not use demographic information such as age, race, or gender as signals to determine the visibility of content or profiles."

Brandeis Marshall, data ethics consultant, told TechCrunch platforms "innately have embedded a white, male, Western-centric viewpoint" due to who trained models. Sarah Dean, Cornell professor, stated "someone's demographics can affect 'both sides' of the algorithm."

We'd like your views:

  • When AI systems demonstrably affect business revenue based on protected characteristics, who bears liability?
  • If systems reward male-stereotyped communication as a proxy for genuine value, is that discrimination?
  • What audit process should precede deploying LLMs in professional opportunity systems?
  • How do you ensure your AI workflows are not biased?

 

4 responses from the Newspage community

Copy all

Star Quote
Copy

Colette Mason
AI Ethics Consultant at Clever Clogs AI
When a platform controlling professional opportunities deploys LLMs trained on decades of male-dominated business content, "we don't use gender as a signal" means nothing if the system learned to treat assertive communication as the gold standard. Small business owners relying on LinkedIn for visibility don't have time to A/B test their gender presentation. They need transparent systems that don't penalise and stereotype. If your AI can't explain why it favours certain voices, it's not fit for purpose in professional infrastructure.
Copy

Editor
Kate Underwood
Founder & Chief People Strategist at Kate Underwood HR and Training
If changing your LinkedIn gender to male makes your business louder overnight, that’s not merit – it’s bias dressed up as tech. Who’s liable? If a UK small business uses AI to hire, decide pay, rate performance or hand out opportunities and women lose out, the employer carries the risk under the Equality Act. “The software did it” won’t protect you. If the system boosts “confident, punchy, matey” language and buries other styles, that can still be discrimination. Same rule. Worse outcome for women. Unless you can prove it’s truly needed for the job, you’re in trouble. Before using LLMs for anything that affects careers: test it on real data, check results for men vs women, write down how it works, keep a human making the final call, and keep checking it every month. If you can’t explain a decision simply, don’t let a machine make it.
Copy

Rohit Parmar-Mistry
Founder at Pattrn Data
If a human manager suppressed women’s visibility by 800%, they’d face a tribunal. Why do we let an algorithm off the hook?

LinkedIn’s defence that they don’t use gender or race as signals is technical sophistry. I’ve spent 10 years in automation, and I know AI is a ruthless pattern-matcher. It finds proxies for value based on historic bias.

I see this personally. To a human, I'm a British expert. To a lazily trained model, the name 'Rohit Parmar-Mistry' implies I'm fresh off the boat, biasing visibility despite the fact I was born here.

When systems impact revenue, liability must sit with the deployer. Hiding behind 'black box' complexity is cowardly. We need adversarial auditing that explicitly tests for these biases before deployment. If your tool silences women or filters out non-Western names 'by accident', you are still responsible for the discrimination you engineered. Innovation without responsibility isn't progress; it is reckless negligence
Copy

Patricia McGirr
Founder at Repossession Rescue Network
When a profile tweak triggers a surge in visibility, the system is shaping outcomes. AI now influences who gets seen and paid, turning bias into a governance and liability issue. Claims about intent or missing demographic data miss the point. If outcomes skew opportunity, commercial exposure follows. Bias enters through data, proxies and feedback loops, then scales fast. Any business relying on these systems carries that risk. The rule is simple: if AI affects commercial outcomes, it requires bias testing, audit and accountability.