Brittani Kaiser, Cambridge AnalyticaIt confirms it Data protection Since he testified to the British Parliament has not improved online 2018 Millions of people may have collected their Facebook data.
Corruption erupted 2016 After coming to light that more than 87 million people may have seen their data on Facebook through personality testing. Although it is not clear how the data was used, Kaiser said that the Cambridge Analytica had done the “job that could face positions” with the election campaign platform. Donald Trump And it Brexit Carried out by leave. Both organizations said there was no agreement signed to work with the analysis company.
“I would like to say,” I would say that there are many parts of the world that are beginning to protect people legally, which is not in 2018, “Kaiser said, ‘Euronaws Next’. He said that there are no federal data laws, even though he claims that there are more data protection laws in Europe (GDPR). America If a person wants to take legal action against companies that use their data without permission, it carrys Long time in court.
“If you don’t want to spend more time in a court or do not handle the law, I will say Technically we are less protected Since technology is the best in choosing objectives, ”he said. Artificial Intelligence (AI) When things are getting worse, especially when the election interference, he said.
“In the days of Cambridge Analytica, we had the most fundamental mechanism of the content used at the time,” he said. “This is not as if it is today, where you can follow the reality AI has been very well“, Said.
Kaiser said he had seen, despite working on protecting digital rights Cheated by pictures created by AI They spread online. On one occasion, he thought something was going on in New York.
“Knowing that it is already difficult to find the real image of something, I would say that it was It is very easy to abuse it for political or commercial purposesOr to want people, “he says.” We are still on the platform that many of our intelligence agencies say that Russia and China and I will spend the same amount of money on misinformation, ”he added.
AI in the election
Last year, more than 60 countries went to the election what world election super bicycle was. Research on whether AI played some of them is low. However, in a recent study of Alan Touring’s Development and Security Technology Center (CETAs), there are 16 viral cases during the United Kingdom’s general elections Misinform Average IA ‘Deepfakes’.
Researchers also analyzed US Elections And examples of false information created by IA were found. Among them were AI’s bots, they followed American voters and followed the allegations against the diaspora, which led to the virus content, thereby obtaining a specific and vast media coverage.
Although there is insufficient evidence that there is insufficient evidence that the US presidential elections have had a firm impact on the results of the US presidential elections, it further states that these contents of AI have affected the US election lecture.Multiply other forms of false information And ignite political debates. “
Apart from elections, Kaiser is one of his biggest concerns, such as governments signing contracts with great technologies dedicated to AI ChaddsOpen. “Since governments are ready to try to use AI, many of them are closed code, black boxes, our data and our government’s data will stop these closed code systems with profit and no security,” he said.
In general, Open source The source code of the software is available for everyone for application, change and distribution, and the data that trains the AI is shared. Instead of, ay Closed code This is material code and AI trained data under the company’s control and property.
However, holding in the hands of anyone who knows how to program large models of open source AI can have impacts on national security. “Open in Open is the majority Government contracts Artificial intelligence, and all government data of all these government companies and sectors that stop our personal data and Open AIA corporate servers are still a big responsibility, ”Kaiser said.
Kaiser added more that this was the case on Cambridge Analytica and Facebook a decade ago, but in the case of Opanai, they are now “Many more dataParticularly Data of the sensitivityIntroduced in these systems. “Kaiser is now pressing, so that governments accept federal and state -wide open source AI companies.
In his opinion, the Open source SE “Essential For the first time, it has been popularly popularized that the public can censor what the government is doing for public public agencies, especially with databases and data. “
AdvertisementIn his opinion, there will be a large number of open source AI systems One step in the protocol As governments begin to accept the AI, it is in the United States that there is no federal laws that protect how they can use their data in these organizations. Kaiser recently accepted a new position in Elisos, an open source AI site that creates AI agents. He is leading the US public sector to create open source AI technology for governments.
Common sense
Kaiser believes it Trump administration Stick More seriously Data security. “This particular government seems to be sure to have a federal technology policy, and appoint technicians and technicians to lead many government departments,” he said. “So I hope that I mean I will see something in the United States,” he added.
“It would be very nice that we can finally see one Federal law To protect American citizens and protect our rights from developing technology. “However, the US made it clear The Summit of Artificial Intelligence in Paris Excessive technical regulation may end in February NoveltyThe country will be at the forefront when referring to the peculiar terms of these technologies.
“Excessive control of the AI sector may end a transformation sector when they depart,” US Vice President JD Wans said. It does not consider Kaiser that it prevents its data protection aspirations in the United States or in the open source, but should be a Balance frame When regulation comes.
Advertisement“Of course, I disagree (the Managing Director of OpenAi) Sam Altman When he says that our models should be allowed to use any data to be competitive. I think it is too fast and I break things for me, “he said.
However, he does not believe that the peculiar regulation will help, because he is not technically implemented, which has happened in some of Europe’s GDP. “But I think one Common senseIf you are writing easily with technicians, this is good for Americans and economy, ”he said.
