The first phase of Adult Sites Against Child Pornography's new system to streamline child porn reporting, a review module in the system they call ASACP Intranet (AI), is now up and running, the group announced September 4.
The group received over 4,000 suspect child porn reports in July and reported over 100 sites to authorities, according to executive director Joan Irvine. She said the AI program means a more efficient way to process the leads the group receives every day, and to help separate actual from merely suspected child porn sites, saving time and resources for the group and for law enforcement agencies..
“AI is a database-driven system that allows investigative information on all leads to be stored for querying and analyzing of the data," said ASACP chief technology officer Brandon Shalton. "AI allows ASACP to post valid child pornography leads to the FBI and NCMEC (and in the near future European hotlines) based on their individual criteria. In addition, ASACP will be able to perform more detailed statistical analysis.”
The review module began operating toward the end of August. “Because of AI, ASACP can continue to send confirmations to people who took their time to report suspect child pornography," Irvine said. "If an e-mail address is included in the report, the individual will receive an email about the results of this review. This feedback loop lets concerned individuals know that the ASACP Leads Reviewer reviewed the submitted reports and educates them on what is not child pornography.”
Once the report is reviewed, Irvine continued, "if it's valid child porn, and we're going to be sending it on to the FBI, we'll let (the individual) know that. If it's not child porn and a model over 18, we let them know that. If it's artistic and therefore deemed not child pornography, we'll let them know that and refer them to our (posting) on what is child pornography. All that, now, is automated in the review module."
The next AI Intranet phase will be a so-called membership module, involving automated search spidering, Irvine said. "Right now, we do everything manually as far as taking a look at the sites, reviewing keywords," she said. The automated spidering is expected to come online in November, she said.
