R.It is the genetic test company 23andme, which carries genetic data from 15 million peopleHe announced bankruptcy on Sunday night after years of financial struggles. This means that all user data is very personal can be offered for sale – and that this wide interest in genetic data can attract attention from artificial intelligence companies that look forward to training their data groups, as experts say.
“The data is the new oil – this is very high -quality oil,” says Sobodha Kumar, Professor of Fox College of Business at Temple University. “With the development of more complex and strict algorithms, this is a golden mine for many companies.”
But any company related to the prosecution, which is trying to get 23andme, will manage great reputation risks. Many people feel terrifying from the thought that they surrendered their genetic data to track their origin, just to be used now in ways they have never agreed.
“Anyone who touches this data is risked,” says Kumar, director of the Business Analysis and Screel Technology Center. “But at the same time, not touching it, they may lose something too much.”
Read more: 23andme was presented for bankruptcy. What does that mean for your account?
LLMS Training
Companies like Openai and Google It has been poured time and resources to make an impact on the medical field, and 23andme data may attract attention from large artificial intelligence companies on financial means to acquire them. 23andme has been estimated all over 48 million dollars This week, a decrease from $ 6 billion in 2021.
These companies seek to build the strongest models of possible general purposes, which are trained in huge amounts of granular data. But the researchers argued that the sources of data are high quality drying UP, which makes new and powerful information sources more disinfection. A techcrunch reconnaissance The owners of investment capital earlier this year found that more than half of the respondents were martyred with “quality or scarcity of their ownership data” as the edge that has the startups of artificial intelligence for their competitors.
“I think it can be a truly valuable data collection for some large artificial intelligence companies because it represents these ground truth data for actual genetic data,” says Kazlauskas. “Some human errors that may exist in vital publications, you can avoid.”
Kumar says that 23ndme data can be of special value for companies to push them to AIC AI, or AIS that can perform tasks without human participation, whether in medical research or decision -making of the company.
“The complete goal of Aicank AI models was a standard approach: you break the smaller parts of the problem and then put them together.”
Google and Openai representatives did not respond immediately to the suspension requests.
The value -based value
23ndme data can also be valuable across different industries using artificial intelligence to sort huge amounts of data – above and above all, medical research.
23ndme has already agreements with Pharmaceutical companies Like Glaxosmithkline, which was used in the company’s data groups in the hope of developing new treatments for the disease. Kumar says that in Temple, he and his colleagues work on a project to create a personal treatment for ovarian cancer patients – and they found that genetic data could be “very strong in understanding the structures that we could not understand,” he says.
However, Alex Chavornkov, founder and director of the Insilico Medicine company, claims that 23andme data may not be valuable as some believe, especially with regard to drug discovery. “Most of the low suspended fruits have already been captured and there is important data in the public field published with the main academic papers,” he wrote in an e -mail until time.
But companies in many other industries are likely to be interested as well. This is an abnormal large and accurate data collection: this amount of genetic data is rarely accessed, especially those that come with personal health and medical records, publicly, says Anna Kazlauskas, CEO of open data laboratories and Vana creator, a network of data owned by users. “All this contextual data makes it of great value – and solid data you get,” she says.
Industries that are likely to be interested in insurance companies, which can use data to identify people with greater health risks, include in order to raise their installments. Financial institutions can follow the relationship between genetic signs and spending patterns in the loan evaluation process. E -commerce companies can use data to customize ads for people with specific medical conditions.
Ethical and privacy concerns
But companies also face great risk of reputation in participation. 23andme suffered from penetration in 2023, which revealed the personal data of millions of users, which greatly hurt the company’s reputation. Kumar says that the offers of offers who come from other industries may have less protection than 23andme data. He says: “What worries me that some companies are not used to having this type of data, and they may not have sufficient governance,” he says.
This is especially dangerous because genetic information is sensitive to its nature and cannot be changed as soon as it is penetrated. The genetic information of the family members of the people who provided their data about the company as well. Given the well -known biases of AI, the misuse of this data can lead to discrimination in areas such as employment, insurance and loans. Friday, California Public Prosecutor Rob Punta Absolute Urgent alert to 23andme customers who advise them to request the company to delete their data and destroy their genetic samples under California privacy law.
Eva Galbenein, director of cybersecurity at the Electronic Borders Corporation, fears that the genetic data of 23andme may exist in a state of permanent flow in the market. “Once the data is sold, there are no limits to the number of times that it may be rested,” she says. This may lead to genetic data in the hands of organizations that may not give priority to ethical considerations or have strong data protection measures.
Zhavoronkov says at ENSILICO that all of these concerns mean that the offers associated with their on their behalf will be bent from trying to buy 23andme and their data. “Their data set is actually poisonous,” he says. “Everyone who buys it and trains it will obtain negative propaganda, and it is possible that the investigator will be investigated or prosecute.”
Regardless of what is happening in the end, Kazlauskas says it is at least grateful because this puzzle has opened larger talks about data sovereignty. She says: “In the future, we must want to avoid this type of situations in which you decide that you want to take a genetic test, and after that after five years, this company fights financially, and it is now presenting your genetic data at the risk of selling it to the highest bidder.” “In this era of artificial intelligence, this data is of great value.”