Data Hygiene and GenAI

By |2024-11-13T10:55:58+00:00November 13th, 2024|

Working in data hygiene, we know a thing or two about the importance of data quality. As more businesses look to harness the power of GenAI, they too are learning that data quality and hygiene are critical to success. AI systems are only as good as the data fed into them, and generative AI (GenAI) is no exception.  BT’s Chief Data & AI Officer, Deepika Adusumilli, recently highlighted that while ideas for GenAI applications are plentiful, turning them into high-impact tools requires rigorous data quality standards. Without clean data, even the best AI strategies struggle—a reminder of the direct mail adage, “rubbish in, rubbish out.” Direct Mail Lessons for GenAI: Data Quality is a Continuous Process The direct mail sector [...]

Ensuring Data Accuracy in the Age of AI: A Lesson from AI-Generated Mistakes from across the Pond

By |2024-08-07T09:55:32+00:00August 7th, 2024|

In recent months, significant incidents have highlighted the pitfalls of AI when it comes to maintaining data accuracy in public records. These events serve as a crucial reminder of the importance of data hygiene, particularly for organisations that rely on precise data processing. The Incident: Pregnant Woman Wrongly Accused One notable incident involved a pregnant woman in Detroit who was falsely arrested based on AI-generated evidence. Porcha Woodruff, eight months pregnant at the time, was wrongly identified by an AI system as a suspect in a carjacking case. This led to her wrongful arrest and detention for 11 hours, causing her significant emotional distress and physical complications leading to hospitalisation. This case is part of a broader pattern, with several [...]

The Curious Case of AI-Generated Data Hacking: A Cautionary Tale

By |2024-02-05T09:30:28+00:00February 5th, 2024|

In the ever-evolving landscape of cybersecurity, a bizarre new trend has emerged, blurring the lines between reality and fiction. Picture this: a hacking forum on a quiet Sunday, buzzing with the advertisement of a stolen data cache from Europcar. The claim? Over 48 million customers' personal information up for sale. The twist? The data might just be a figment of artificial intelligence's imagination, specifically conjured up by ChatGPT. Europcar's response was swift, with spokesperson Vincent Vevaud debunking the claim after a meticulous examination. The inconsistencies were glaring: The purported number of records didn't match Europcar's database. Sample data was likely AI-generated, featuring non-existent addresses, mismatched ZIP codes, and bizarre email domains. None of the email addresses were found in Europcar's [...]

Go to Top