My Lords, I echo the noble Lord’s words. We also welcome these amendments. As has been said, this issue was raised by the academic community, whose primary concern was that the way the Bill had originally been phrased would make important security research illegal and weaken data protection for everyone by that process. It would also mean that good and valid research going on in our high-quality institutions might be at risk.
I do not in any sense want to question the amendments’ approach, but I have been in further correspondence with academics who have asked us to make a few points. I am looking for a sense that the issues raised are being dealt with. Either a letter or a confirmation that these will be picked up later in the process of the Bill is all that is necessary.
First, it is fairly common-sense to say that companies probably would not be very happy if a researcher picks up that they are not doing what they say on the tin—in other words, if their claim that their data has been anonymised turns out not to be the case. Therefore, proposed new subsection (2)(b) may well be used against researchers to threaten or shut down their work. The wording refers to “distress” that might be caused, but,
“without intending to cause, or threaten to cause, damage or distress to a person”,
seems a particularly weak formulation. If it is only a question of distress, I could be distressed by something quite different from what might distress the noble Lord, who may be more robust about such matters. I think that is a point to take away.
Secondly, we still do not have, despite the way the Minister introduced the amendment, definitions in the Bill that will work in law. “Re-identification”, which is used in the description and is part of the argument around it, is still not defined. Therefore, in proposed new Clause 161A(3), as mentioned by the noble Lord who introduced the amendment, the person who,
“notified the Commissioner or the controller responsible for de-identifying the personal data about the re-identification”,
has to do this,
“without undue delay, and … where feasible, not later than 72 hours after becoming aware of it”.
That is a very tight timetable. Again, I wonder if there might be a bit more elasticity around that. It does say “where feasible”, but it puts rather tight cordon around that.
We are trying to make it safe for researchers and data scientists to report improperly de-identified data, but in the present arrangements the responsibility for doing all this lies with the researcher. We are asking a researcher to go to court, perhaps, and defend themselves, including arguing that they have satisfied Clause 162(2)(a) and (b) and Clause 162(3)(a), (b) and (c), which is a fairly high burden. All in all, we just wonder whether how this has been framed does the trick satisfactorily. I would be grateful for further correspondence with the Minister on this point.
Finally, there is nothing in this amendment about industry. It may not be necessary but it raises a question that has been picked up by a couple of people who have corresponded with us. The burden, again, is on the researcher. Is there not also a need to try to inculcate a culture of transparency in the anonymisation processes which are being carried out in industry? In other words, if there is a duty on researchers to behave properly and do certain things at a certain time, should there not also be a parallel responsibility, for example, on companies to properly and transparently anonymise the data? If there is no duty for them to do it properly, what is in it for them? It may well be that that is just a natural aspect of the work they are doing, but maybe the Government should reflect on whether they are leaving this a little one-sided. I put that to the Minister and hope to get a response in due course.