"The report shows that 91% of organizations believe sensitive data should be allowed in AI model training. At the same time, 78% report high concern about theft or breaches. Experts say this discrepancy stems from a lack of understanding about the permanence of data in AI systems. Once sensitive information is used to train a model, it cannot be fully removed or made completely secure. This creates a lasting exposure risk, particularly when personal or confidential data is involved."
"Steve Karam, Principal Product Manager at Perforce, noted that organizations face dual pressures: the need to innovate rapidly with AI while ensuring that privacy and compliance standards are met. He emphasized that personally identifiable information (PII) should never be used in model training, and that alternative approaches, such as synthetic data, can provide secure pathways for AI development. This paradox is not merely theoretical. Organizations often underestimate the ways in which AI systems retain and propagate training data."
Organizations overwhelmingly support using sensitive data in AI model training, with 91% endorsing its use. Simultaneously, 78% of organizations express high concern about theft or breaches. Experts attribute this paradox to limited understanding of data permanence in AI systems; data used for training cannot be fully removed or made completely secure, creating lasting exposure risks for personal and confidential information. Organizations face pressure to innovate with AI while meeting privacy and compliance standards. Personally identifiable information (PII) should not be used in model training. Synthetic data and other alternative approaches can enable secure AI development. AI model outputs can inadvertently expose sensitive information even in controlled environments.
Read at App Developer Magazine
Unable to calculate read time
Collection
[
|
...
]