Google Reverses AI Health Data Policy After Employee Backlash
Google initially required employees to share health data with a third-party AI tool called Nayya to receive health benefits, but reversed the policy after employee pushback and media reporting. The company clarified that participation is optional and employees can maintain benefits eligibility without sharing data.
What happened
Google's internal memo told US employees they must opt into Nayya, a third-party AI healthcare tool, to receive health benefits during the upcoming enrollment period. According to internal guidelines, staff who declined to opt for the tool would not be eligible for any health coverage, with data sharing described as mandatory under Alphabet's benefits framework.
After Business Insider reported on the policy and employee backlash, Google changed course and told employees that opting into Nayya is voluntary and that they can maintain benefits eligibility without sharing data. A Google spokesperson stated, 'Our intent was not reflected in the language on our HR site. We've clarified it to make clear that employees can choose to not share data, without any effect on their benefits enrollment.'
Reported impact
Employees raised privacy concerns and posted messages on company forums, Q&A websites and message boards calling the policy 'coercive and dark pattern.' One employee wrote, 'Consent for an optional feature like benefits usage optimization is not meaningful if it's coupled to a must-have feature like Google's HEALTH PLANS.'
Index analysis
Nayya's platform uses artificial intelligence to offer personalized benefits recommendations based on employee health and lifestyle data, including sensitive information such as pay, gender, and social security number.
Governance controls
Google updated its internal benefits site guidelines after the policy backlash to clarify that employees can choose not to share data with Nayya without affecting their benefits enrollment or health plan eligibility.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.