The attribution of cell phones to indicted suspects in the 2005 assassination of former Prime Minister Rafik Hariri was the central focus of the Special Tribunal for Lebanon Tuesday. Prosecution analyst Andrew Donaldson returned to the witness stand, where he was cross-examined by David Young, lead counsel for the defense of Assad Sabra, one of the four accused of involvement in the bomb attack that killed Hariri and 21 others.
The defense criticized Donaldson’s failure to define his own role once the decision was made in 2012 that he would be used as an attribution witness. Donaldson produced numerous reports for the prosecution during that time, but failed to address how he defined “attribution,” drawing criticism from both the counsel for the defense, and the trial judges. This problem was particularly urgent, one critic argued, given the fact that this is one of the world’s most significant cases to be relying heavily on cellphone data.
When the defense counsel suggested that Donaldson was “leaving it a bit late to explain what you mean by attribution,” the expert witness responded that such a definition had been outside of his remit.
“I don’t get to determine the reports I produce,” he said. “I wasn’t asked to do it ... you would need to ask [the prosecution] why they never asked me to [define the term].”
Emphasizing the importance of cellphone data to the case, Trial Chamber President Judge David Re followed up the point. “Without attribution there is no case against any of the accused,” he said. “I think you’re conceding in hindsight that stating what you mean by attribution might have been helpful for the reader in the reports.”
Donaldson told the tribunal that there was a spectrum regarding the quality of attribution, noting that in the United Kingdom a grading system is used, according to which mobile phone data is the “only intelligence source that could be graded A1 – in other words the best information.” This, he said, was due to the imperviousness of such evidence to bias and memory loss many years after the events in question.
Donaldson agreed with the defense counsel that cellphone data attribution was not an exact science.
“Anyone can send any message. Anything is possible. It just comes down to probabilities.”
He discussed the many different factors involved in attributing cellphone data, pointing out, for example, that a single cellphone may have multiple users. Documentary evidence such as a phone subscription or a passport application for a telephone contract was not given significant weight, he said, as the phone may subsequently change hands.
With regards to SMS messages, Donaldson said that it was not the volume of such messages that was of primary importance, but the content, which might prove far more useful in establishing a relationship to another person or placing the user in a specific location.
The counsel for the defense asked whether there might be factors that could lead to a misattribution of a device. Donaldson answered in the affirmative – suggesting factors such as “bad working practices, or making assumptions and drawing inferences without testing them.”
He refuted that his team might have been prone to this level of human error. “People on my team don’t make those mistakes,” he said.