Meta iends to use artificial ielligence -based systems to assess the risk of its products and abandon the use of traditional systems.
According to iernal documes that NPR media has observed, the meta seeks to evaluate artificial ielligence -equipped systems responsible for evaluating poteial damage and privacy risks in the company’s apps, such as Instagram and WhatsApp. The system evaluates up to 5 % of the updates made in meta services.
Artificial Ielligence is responsible for assessing the risk of meta products
The Meta and the Federal Commerce Commission of the Year has obliged the company to review its products in terms of privacy and evaluate the possible risks of each update. To date, these reviews were carried out by manpower, but artificial ielligence will soon be in charge.
In the new system, Meta product teams are apparely asked to complete a questionnaire about their activities. These teams will come up with a “immediate decision” proposal. The offer includes ideified risks and methods of removing them. Product teams must meet these prerequisites before launching their product.
This artificial ielligence -based approach allows Meta to update its products faster, but a former company executives told NPR that it has more risks; Because it may be possible to preve the negative effects of product changes before it causes problems in the real world.
A spokesman for Meta said in a stateme that the company has invested more than $ 5 billion in its privacy program and is committed to providing innovative products to the public while adhering to legal obligations. “We use technology to create coordination and predictability in low -risk decisions, and we rely on human expertise for accurate evaluations and monitoring new or complex issues,” he said.
