[AI Minor News Flash] NYC Public Hospitals to Cut Ties with Palantir! Shifting to In-House Systems Over Data Privacy Concerns

📰 News Overview

• NYC Health + Hospitals will terminate its contract with Palantir, which has cost about $4 million since November 2023, in October.
• The hospital system has announced plans to transition completely to an in-house system, no longer using Palantir’s applications.
• Palantir also holds a massive contract with the UK’s National Health Service (NHS), but criticism is mounting in both the US and UK regarding data privacy and the risks of “data re-identification.”

💡 Key Points

• The contract included “revenue cycle optimization,” aimed at reviewing patient health records and maximizing claims for public benefits like Medicaid.
• Controversially, there was a clause allowing Palantir to “de-identify” patient data with hospital consent, enabling its use for non-research purposes.
• Privacy experts warn that Palantir’s access to extensive government records raises the potential risk of re-identifying individuals from anonymized data.

🦈 Shark’s Eye (Curator’s Perspective)

Implementing AI for revenue improvement is efficient, but this contract termination signals a move towards reclaiming “data sovereignty” for the hospitals! The clause allowing data use for non-research purposes is particularly concerning—there’s plenty to bite into here! Palantir is constantly under scrutiny due to accusations of data sharing with ICE, and New York City’s choice to go “in-house” could set a precedent impacting other public agencies! This is a classic case where the lack of trust led to a contract cancellation rather than the specifics of implementation.

🚀 What’s Next?

To avoid reliance on Palantir, other large public agencies may also pursue in-house AI systems or switch to vendors with greater transparency. Meanwhile, the decision in NYC could pose further challenges for Palantir’s plans within the NHS in the UK.

💬 Haru-Same’s Take

If you’re swimming in a sea of data, you better know who you’re feeding (the data) to! Going in-house is tough, but nothing beats peace of mind! 🦈🔥

📚 Glossary

• De-identification: The process of removing personally identifiable information from data sets, such as names or social security numbers.

• Revenue Cycle Optimization: Managing and improving how hospitals receive proper compensation for services provided, ensuring no claims are overlooked by insurers and public entities.

• NHS Federated Data Platform (FDP): A foundational system being developed by the UK’s National Health Service to integrate and analyze data across regions.

• Source: New York City hospitals drop Palantir as controversial AI firm expands in UK