Lawmakers who want to address privacy challenges posed by the burgeoning artificial intelligence industry should start by tackling so-called “commercial data surveillance.”
That's according to Chris Lewis, president and CEO of the advocacy group Public Knowledge, who testified Wednesday at a Senate AI Insight Forum on privacy and liability issues.
“Right now the largest tech companies have a monopoly not only on computational resources, but personal data,” Lewis said in his written testimony. “The current AI boom is primarily powered by commercial data surveillance, which has resulted in a small group of companies establishing dominance over the digital world.”
Lewis specifically touted the American Data Privacy and Protection Act -- a federal privacy bill introduced last year that, if passed, would have imposed sweeping curbs on data collection and use.
He wrote that the bill “would minimize the amount of personal data collected, would give people rights over their data, encourage competition, and integrate important civil rights protection.”
While the bill included numerous restrictions, one of the most striking was that it would have outlawed some forms of behavioral advertising. Specifically, the version of the bill discussed last year would have prohibited companies from collecting or processing data about web users' online activity across sites and over time for ad purposes.
Ad groups opposed the bill, with Interactive Advertising Bureau CEO David Cohen going so far as to say it would “destroy our industry.”
The House Energy and Commerce Committee advanced the bipartisan bill by a vote of 53-2 in July 2022, but the full House didn't vote on the measure, and the Senate didn't hold hearings on it.
The nonprofit Mozilla, which develops the Firefox browser, likewise called for new privacy protections.
“What we need to prevent is a race to the bottom when it comes to privacy -- so passing a federal privacy law, setting effective rules of the road, is paramount,” the organization stated, adding that it endorsed the American Data Privacy and Protection Act last year.
The watchdog Center for Democracy & Technology expressed support for laws that would prohibit companies from collecting or processing more data than needed to provide a particular service or product.
“Data collection has run rampant in the digital age in large measure because companies have economic incentives to amass large pools of data, such as using data to target advertising based on people’s behavior,” Samir Jain, vice president of policy at that group, said in his written testimony. “The need for large datasets to train AI systems provides yet another reason for companies to collect or repurpose extensive data about everyone online.”
While ad industry representative Stu Ingis, chairman of the law firm Venable, reiterated that industry organizations support a federal privacy law, he also stressed that artificial intelligence requires data.
“AI is dependent on data. It is core to creating AI models, to machine learning, and -- often -- AI is used to sift through large amounts of data to create predictions, find anomalies, and detect threats,” he said in his written statement.
He added that privacy “should be considered throughout AI model development and implementation” and that the “confidentiality and sensitivity of input and output data should be considered, especially in contexts where AI-based determinations impact rights and eligibility for certain opportunities.”
Senate Majority Leader Chuck Schumer (D-New York), who convened the forum, stated in his opening remarks that artificial intelligence and data are already “inextricably linked.”
“We see it everywhere today, from social media to advertising, AI is already using our data to curate what we see on the internet,” he stated. “But advances in this cutting-edge technology could allow for much more problematic uses of AI, such as emotional manipulation.”