Keeping personal information private in the digital era requires more than the simple intention to do the right thing, as the New South Wales Medical Council found out last week.
The statutory body’s staff made sure to black out a person’s name it was legally obliged not to publish before uploading a PDF document, without realising the person’s identity remained in underlying data that would link them to the document in search engine results.
The Medical Council’s then legal director Miranda St Hill had to contact the Australian arm of Google before the problem could be fixed, and even then it was not a simple process to take back what was inadvertently put out there.
The mistake meant the council breached the privacy of a doctor and her son for a five-month period in 2011, NSW Civil and Administrative Tribunal senior member Naida Isenberg ruled last week.
The doctor was involved in a series of medical registration appeals heard by the NCAT’s Medical Tribunal and successfully applied for a non-publication order in respect to her identity at the conclusion of the case. The council made the error when it published that decision online:
“Ms St Hill explained that the method used to ‘redact’ the Applicant’s name merely placed an opaque ‘block’ over the top of the ‘redacted’ words. While this prevented the human eye from reading the Applicant’s name, it allowed, as was later discovered, Google webcrawlers (also called ‘Googlebots’) to read this information and to link the publication to a search of ‘Dr [AIN]’ (and similar).
” … the fact that a Google search of the Applicant’s name resulted in a link to the decision meant it could readily be inferred that the decision related to her.”
Her son was also identifiable because the fact she had taken time off to care for him was an element of the previous case. The council only realised Google’s indexing software had found the name that remained buried in the offending PDF when it was informed by the doctor’s legal representatives several months after it published it.
The doctor was then further dismayed after being told the data had been fully removed from the Medical Council website four days before the change had actually taken effect, and the council “needed to be told on several occasions that the publication continued to be available via a Google search of the applicant’s name before it was successfully removed”.
Once the complex and clever indexing programs behind the scenes of web searches grab hold of something, removing all trace of it is not simple and takes time.
The link between the name and the NCAT judgement was still there for anyone to see more than two weeks after St Hill first contacted Google Australia for help.
A month late, Google Australia asked for the legal documents and, according to Isenberg, received a copy of the decision “but apparently not the NPO, because no formal Order had been obtained”. The problem seems to have finally been fixed about six months after the initial mistake but the hearing did not establish exactly what compelled Google to action in the end:
“Unsurprisingly, on 20 July 2011, Google Australia informed the Respondent that it would take no action.
“Nonetheless, on 20 July 2011, when Ms St Hill arranged for a Google search of the Applicant’s name, no link to the Medical Tribunal’s decision appeared in the first 100 returned searches.”
The problem didn’t go away completely. Searching for the doctor’s name still brings forth a link to the Medical Council’s website, but at least her name appears nowhere on the site and the nature of her relationship to the body is not clear. That’s not a breach of NSW privacy law, and the tribunal accepted any further breach of the original non-publication order would not be the Medical Council’s fault.
Any feedback or news tips? Here’s where to contact the relevant team.