1 00:00:06,440 --> 00:00:09,600 Speaker 1: Kilda. I'm Chelsea Daniels and this is the Front Page, 2 00:00:10,000 --> 00:00:18,280 Speaker 1: a daily podcast presented by the New Zealand Herald. Sending 3 00:00:18,320 --> 00:00:21,480 Speaker 1: your DNA to a website to find out your background 4 00:00:21,520 --> 00:00:24,680 Speaker 1: has become a trendy thing to do. But is there 5 00:00:24,720 --> 00:00:28,320 Speaker 1: a hidden cost to this? DNA testing company twenty three 6 00:00:28,360 --> 00:00:32,320 Speaker 1: and meters is in a financial crisis, raising questions about 7 00:00:32,320 --> 00:00:35,400 Speaker 1: what will happen to the data it holds on fifteen 8 00:00:35,680 --> 00:00:40,760 Speaker 1: million customers worldwide. New Zealanders are concerned their genetic information 9 00:00:40,960 --> 00:00:44,839 Speaker 1: could be on sold and used for other purposes, including 10 00:00:44,840 --> 00:00:48,880 Speaker 1: by insurance companies or law enforcement. Today on the Front Page, 11 00:00:49,120 --> 00:00:57,160 Speaker 1: privacy lawyer Rich Shearer joins us to discuss can you 12 00:00:57,200 --> 00:01:01,280 Speaker 1: tell us how these DNA services work? What does it entail? 13 00:01:01,640 --> 00:01:04,600 Speaker 2: What happens is that you usually you spit in a 14 00:01:04,640 --> 00:01:07,400 Speaker 2: little tube and you send it off to them, and 15 00:01:07,480 --> 00:01:12,240 Speaker 2: then they run their DNA sequencing over it to extract 16 00:01:12,280 --> 00:01:15,080 Speaker 2: the DNA sequence from it, and then they store that 17 00:01:15,440 --> 00:01:19,000 Speaker 2: DNA sequence as data on their systems. I think sometimes 18 00:01:19,000 --> 00:01:21,759 Speaker 2: they keep the sample itself, sometimes they get rid of it, 19 00:01:21,800 --> 00:01:24,080 Speaker 2: depends whether they want to keep it for future use 20 00:01:24,200 --> 00:01:25,920 Speaker 2: or whatever they want to do with it. But generally, 21 00:01:25,959 --> 00:01:28,679 Speaker 2: it's that it's not so much the physical sample that 22 00:01:28,760 --> 00:01:32,640 Speaker 2: I'm focused on, it's more the DNA sequence, the data 23 00:01:32,720 --> 00:01:35,800 Speaker 2: that is generated from that and stored by that organization. 24 00:01:36,040 --> 00:01:40,160 Speaker 2: What they then do is make it available to other 25 00:01:40,560 --> 00:01:45,000 Speaker 2: organizations for those organizations to use. They would, in theory 26 00:01:45,040 --> 00:01:47,680 Speaker 2: anonymize all of the DNA so that you have a 27 00:01:47,760 --> 00:01:52,040 Speaker 2: DNA sequence that isn't necessarily labeled with Chelsea Daniels or 28 00:01:52,120 --> 00:01:54,919 Speaker 2: Rick Scherer or anything else, and they put it together 29 00:01:54,960 --> 00:01:59,000 Speaker 2: with thousands of others and provide that to a research 30 00:01:59,080 --> 00:02:02,000 Speaker 2: organization or a drug company or something like that who 31 00:02:02,080 --> 00:02:04,720 Speaker 2: might be using it to say, well, we want to 32 00:02:04,880 --> 00:02:09,240 Speaker 2: assess whether we can detect certain sequences might give rise 33 00:02:09,320 --> 00:02:12,960 Speaker 2: to certain characteristics of a person, and therefore we can 34 00:02:13,000 --> 00:02:15,560 Speaker 2: design drugs, or we can predict disease and so on 35 00:02:15,560 --> 00:02:17,240 Speaker 2: and so forth. So there's a lot of research that 36 00:02:17,280 --> 00:02:20,639 Speaker 2: goes on around it. There's also been potential use by 37 00:02:20,960 --> 00:02:25,040 Speaker 2: law enforcement agencies of some of this material, not necessarily 38 00:02:25,080 --> 00:02:30,200 Speaker 2: because the organizations give it to law enforcement agencies voluntarily, 39 00:02:30,280 --> 00:02:33,000 Speaker 2: but the law enforcement agency might come along, so, look, 40 00:02:33,040 --> 00:02:37,200 Speaker 2: we understand that you've got DNA for a person, we're 41 00:02:37,240 --> 00:02:40,760 Speaker 2: interested in them as potentially as a subject of an investigation, 42 00:02:41,120 --> 00:02:45,040 Speaker 2: or indeed to match up against another person who's DNA. 43 00:02:45,120 --> 00:02:47,640 Speaker 2: We have to determine whether, in fact that is the 44 00:02:47,680 --> 00:02:49,639 Speaker 2: person that we're interested in. So there's a number of 45 00:02:49,680 --> 00:02:52,000 Speaker 2: different uses which are being made at the moment of 46 00:02:52,000 --> 00:02:52,880 Speaker 2: the sort of material. 47 00:02:54,720 --> 00:02:55,320 Speaker 3: I did that too. 48 00:02:55,440 --> 00:02:57,440 Speaker 4: I sent my spirit into one of those websites, and 49 00:02:57,480 --> 00:03:00,520 Speaker 4: it turns out of sixteen percent Irish and ten percent 50 00:03:00,600 --> 00:03:02,360 Speaker 4: Hungarian and the rest is Italian. 51 00:03:02,520 --> 00:03:02,840 Speaker 3: That's me. 52 00:03:03,639 --> 00:03:05,240 Speaker 4: I did it, and I found out, guys, I found 53 00:03:05,240 --> 00:03:07,640 Speaker 4: out I'm seventy five percent French. 54 00:03:08,120 --> 00:03:09,200 Speaker 2: I don't even speak French. 55 00:03:09,440 --> 00:03:13,680 Speaker 3: Well, oh, I sent my name and I found out 56 00:03:13,680 --> 00:03:18,160 Speaker 3: that I am, let's see here, ninety two percent dumbass 57 00:03:18,200 --> 00:03:20,680 Speaker 3: for sending a DNA sample to a random website that 58 00:03:20,720 --> 00:03:24,120 Speaker 3: now has full copyright and unregulated access to my genetic 59 00:03:24,160 --> 00:03:25,160 Speaker 3: code for the rest of time. 60 00:03:28,800 --> 00:03:32,720 Speaker 1: So we're talking about sites like ancestry dot com twenty 61 00:03:32,760 --> 00:03:34,960 Speaker 1: three and MENI family Tree and that kind of thing. 62 00:03:35,160 --> 00:03:37,240 Speaker 2: Right, Yeah, that's right. And I mean, you know, there's 63 00:03:37,280 --> 00:03:41,600 Speaker 2: a huge business worldwide obviously in genealogy, and this is 64 00:03:41,720 --> 00:03:45,200 Speaker 2: part of that in some senses, because the business model 65 00:03:45,320 --> 00:03:49,120 Speaker 2: for these organis these companies is get your information in 66 00:03:49,200 --> 00:03:52,120 Speaker 2: and then they monetize it by making development of others 67 00:03:52,160 --> 00:03:54,400 Speaker 2: at cost. You pay a little bit. I can't remember 68 00:03:54,440 --> 00:03:57,080 Speaker 2: what people pay to actually do to give people these things, 69 00:03:57,080 --> 00:03:59,160 Speaker 2: but it's not very much. And that's not where they 70 00:03:59,200 --> 00:04:01,480 Speaker 2: make a lot of them. They make their money out 71 00:04:01,520 --> 00:04:04,640 Speaker 2: of making it available to other organizations. But from the 72 00:04:04,680 --> 00:04:07,680 Speaker 2: individual's point of view and giving it to these organizations, 73 00:04:07,720 --> 00:04:11,160 Speaker 2: it's you know, obviously it's genealogy is really interesting. You know, 74 00:04:11,240 --> 00:04:14,520 Speaker 2: you might find another family member, or you might find 75 00:04:14,520 --> 00:04:17,600 Speaker 2: some characteristics of your DNA that you weren't aware of, 76 00:04:17,760 --> 00:04:20,720 Speaker 2: so it's very interesting. I can well understand why people 77 00:04:20,760 --> 00:04:23,119 Speaker 2: do it. That's why there are TV programs about people 78 00:04:23,160 --> 00:04:25,560 Speaker 2: trying to find origins that they're always right. 79 00:04:25,640 --> 00:04:28,360 Speaker 1: Quite highly Now, it seems like a fun thing to do. Hey, 80 00:04:28,400 --> 00:04:30,640 Speaker 1: spit in a tube, send it off and find out 81 00:04:30,640 --> 00:04:33,039 Speaker 1: a few weeks later you're sixty three percent Scottish. But 82 00:04:33,120 --> 00:04:36,960 Speaker 1: when you send your DNA into one of these services, 83 00:04:37,400 --> 00:04:41,039 Speaker 1: what kind of privacy laws are you rescinding? 84 00:04:41,520 --> 00:04:45,919 Speaker 2: When you do this, you free to their terms of 85 00:04:46,040 --> 00:04:50,240 Speaker 2: use and to their privacy policies. By and large, these 86 00:04:50,360 --> 00:04:54,040 Speaker 2: organizations are based in the United States, and so whenever 87 00:04:54,040 --> 00:04:57,120 Speaker 2: you agree to give any information to anybody, whether it's 88 00:04:57,200 --> 00:05:00,000 Speaker 2: using Facebook or going online or whatever, you are saying, 89 00:05:00,320 --> 00:05:03,760 Speaker 2: up to privacy policies, I guess that's the first gateway 90 00:05:03,800 --> 00:05:06,520 Speaker 2: that you go through as the privacy policy. Then there 91 00:05:06,520 --> 00:05:09,960 Speaker 2: are laws which might apply over and above the privacy 92 00:05:09,960 --> 00:05:14,080 Speaker 2: policy in the particular jurisdiction or country where the organization 93 00:05:14,320 --> 00:05:17,839 Speaker 2: is based. So in these instances you've got a privacy 94 00:05:17,839 --> 00:05:21,480 Speaker 2: policy which we can read. Unfortunately, as a person who 95 00:05:21,560 --> 00:05:24,000 Speaker 2: drafts these privacy policies, I have a great deal of 96 00:05:24,000 --> 00:05:26,080 Speaker 2: sympathy for people who think they want to try and 97 00:05:26,120 --> 00:05:28,920 Speaker 2: read them, because they're a good bedtime reading, but only 98 00:05:28,960 --> 00:05:31,640 Speaker 2: for lawyers. So you know, reading thirty five pages of 99 00:05:31,640 --> 00:05:34,359 Speaker 2: a privacy policy to try and discern exactly what the 100 00:05:34,480 --> 00:05:37,400 Speaker 2: organization is going to do with my personal information is 101 00:05:37,400 --> 00:05:40,120 Speaker 2: not something that I can recommend to anyone in their alone, 102 00:05:40,320 --> 00:05:43,160 Speaker 2: non lawyers. So there's the privacy policy, then there might 103 00:05:43,200 --> 00:05:45,960 Speaker 2: be the law in this case in the United States. Now, 104 00:05:46,000 --> 00:05:48,680 Speaker 2: the United States, it's the name suggests, but we tend 105 00:05:48,680 --> 00:05:50,839 Speaker 2: to forget, is made up of a lot of states. 106 00:05:51,080 --> 00:05:55,120 Speaker 2: The United States doesn't have an overarching privacy a federal 107 00:05:55,440 --> 00:05:59,359 Speaker 2: privacy law which applies across everybody in the United States 108 00:06:00,240 --> 00:06:03,080 Speaker 2: states that have passed them. Each law is different, so 109 00:06:04,160 --> 00:06:07,480 Speaker 2: some of them would have cover the certain aspects, others won't, 110 00:06:07,760 --> 00:06:10,599 Speaker 2: depending on what the information has been collected, how many 111 00:06:10,600 --> 00:06:12,400 Speaker 2: people they are collecting it from, and what they do 112 00:06:12,520 --> 00:06:14,760 Speaker 2: with it. So it is a very much a patchwork 113 00:06:15,120 --> 00:06:18,599 Speaker 2: of privacy protection. Obviously, as I said, you've got a 114 00:06:18,600 --> 00:06:21,680 Speaker 2: privacy policy. The American way of doing privacy policies is 115 00:06:21,760 --> 00:06:25,200 Speaker 2: very much you give us consent to do X, y 116 00:06:25,360 --> 00:06:28,080 Speaker 2: Z with the information that you're giving to us. And 117 00:06:28,200 --> 00:06:32,480 Speaker 2: so these policies do tend to say we can provide 118 00:06:32,560 --> 00:06:36,440 Speaker 2: your personal information to other people for these purposes, research, 119 00:06:36,440 --> 00:06:40,359 Speaker 2: et cetera. Because it becomes that that database of DNA 120 00:06:40,720 --> 00:06:44,559 Speaker 2: becomes an asset of the organization. They've created the data 121 00:06:44,680 --> 00:06:47,400 Speaker 2: out of your DNA. You have rights in respect of 122 00:06:47,400 --> 00:06:51,279 Speaker 2: your DNA yourself, but the data, the DNA sequence that's 123 00:06:51,320 --> 00:06:55,680 Speaker 2: created is not yours. It is data digital information, which 124 00:06:55,720 --> 00:06:59,680 Speaker 2: is created by the organization, so they own that. And 125 00:06:59,760 --> 00:07:03,360 Speaker 2: so the privacy policies will generally often say, look, in 126 00:07:03,440 --> 00:07:06,080 Speaker 2: the event of a sale of our business, we can 127 00:07:06,240 --> 00:07:09,960 Speaker 2: sell our assets, including that personal information you've given. Now 128 00:07:10,280 --> 00:07:13,640 Speaker 2: there maybe some controls around that, not necessarily the level 129 00:07:13,640 --> 00:07:15,960 Speaker 2: of control that you or I might want to have 130 00:07:16,120 --> 00:07:19,360 Speaker 2: in respect of very personal information like DNA. Twenty three 131 00:07:19,440 --> 00:07:21,600 Speaker 2: and meters on the verge of getting delisted now from 132 00:07:21,600 --> 00:07:22,160 Speaker 2: the Nasdaq. 133 00:07:22,760 --> 00:07:25,400 Speaker 3: Drama has been unfolding this week. It's only set to 134 00:07:25,400 --> 00:07:25,960 Speaker 3: wrap up. 135 00:07:26,240 --> 00:07:29,040 Speaker 4: Twenty three Meters built a multi billion dollar brand off 136 00:07:29,080 --> 00:07:32,200 Speaker 4: of collecting people's spit. The idea was that people could 137 00:07:32,240 --> 00:07:34,760 Speaker 4: take control of their health by learning about their genetics, 138 00:07:35,160 --> 00:07:38,679 Speaker 4: and it took off. But just two years after going public, 139 00:07:38,800 --> 00:07:41,560 Speaker 4: the company went from being valued at six billion dollars 140 00:07:41,680 --> 00:07:44,680 Speaker 4: to treating below one dollar. Now, twenty three and Meter 141 00:07:44,800 --> 00:07:47,600 Speaker 4: is rapidly approaching the November deadline to propose an action 142 00:07:47,720 --> 00:07:50,320 Speaker 4: plan or be delisted from the Nasdaq. 143 00:07:50,960 --> 00:07:54,440 Speaker 1: What are the concerns around twenty three and ME going 144 00:07:54,640 --> 00:07:57,360 Speaker 1: under and what happens to the rights of the DNA 145 00:07:57,480 --> 00:07:59,119 Speaker 1: records that they have on file. 146 00:07:59,360 --> 00:08:03,200 Speaker 2: They haven't gone yet. They're in financial difficulty. Now there 147 00:08:03,240 --> 00:08:05,960 Speaker 2: could be I guess two results of that or three results. 148 00:08:05,960 --> 00:08:08,640 Speaker 2: I suppose they could come out of financial difficulty, they 149 00:08:08,640 --> 00:08:11,119 Speaker 2: could get more investment, and suddenly they take off again, 150 00:08:11,200 --> 00:08:13,920 Speaker 2: so that would be all good. They could decide, look, 151 00:08:14,000 --> 00:08:17,560 Speaker 2: our shareholders are not putting any money more money into us. 152 00:08:18,120 --> 00:08:20,440 Speaker 2: Our share price is going down as it is at 153 00:08:20,480 --> 00:08:22,680 Speaker 2: the moment, and therefore, in order to act in the 154 00:08:22,680 --> 00:08:25,480 Speaker 2: best interests of our shareholders, we should sell our business 155 00:08:25,600 --> 00:08:28,239 Speaker 2: to someone else. In that business will include the assets, 156 00:08:28,280 --> 00:08:31,600 Speaker 2: which will include the DNA data set, So that could 157 00:08:31,600 --> 00:08:34,640 Speaker 2: be a scenario or of West Coast scenario. The company 158 00:08:34,679 --> 00:08:37,360 Speaker 2: can't do that and it can no longer continue and 159 00:08:37,440 --> 00:08:41,520 Speaker 2: it goes bankrupt, and a bankruptcy liquidator that would be 160 00:08:41,559 --> 00:08:44,000 Speaker 2: in New Zealand that is appointed and they then sell 161 00:08:44,040 --> 00:08:46,560 Speaker 2: the assets. Of course, at that stage all bets are off. 162 00:08:46,600 --> 00:08:48,520 Speaker 2: They would sell to whoever they can get the best 163 00:08:48,559 --> 00:08:51,640 Speaker 2: price for those assets, which could be anyone. So in 164 00:08:51,679 --> 00:08:54,200 Speaker 2: any of those scenarios where the assets are sold, we 165 00:08:54,240 --> 00:08:57,400 Speaker 2: don't have any control over who might be the purchaser 166 00:08:57,600 --> 00:09:00,319 Speaker 2: of those assets in what they then might use stand 167 00:09:00,320 --> 00:09:04,120 Speaker 2: information for whether they are bound by the privacy policy 168 00:09:04,280 --> 00:09:08,080 Speaker 2: that is in place is questionable. In most privacy policies 169 00:09:08,240 --> 00:09:11,040 Speaker 2: will say look, we can change this policy whenever we 170 00:09:11,200 --> 00:09:12,760 Speaker 2: like by telling you, and if you don't want to 171 00:09:12,840 --> 00:09:15,440 Speaker 2: use the service after that, then stop using it. But 172 00:09:15,480 --> 00:09:18,480 Speaker 2: if you do continue to use it after we've notified 173 00:09:18,520 --> 00:09:20,640 Speaker 2: you of the change, then you're bound by the new policy. 174 00:09:20,720 --> 00:09:22,839 Speaker 2: So it could be that there's a change in that area. 175 00:09:23,040 --> 00:09:26,800 Speaker 2: Some states in the United States might restrict that, but 176 00:09:26,960 --> 00:09:28,920 Speaker 2: as I say, it's a bit patchwork in the US 177 00:09:28,960 --> 00:09:31,600 Speaker 2: as to how that would apply. You know, West caase scenario. 178 00:09:31,880 --> 00:09:35,120 Speaker 2: This information could end up with a company that wants 179 00:09:35,160 --> 00:09:38,640 Speaker 2: to do something completely different with it, wants to target 180 00:09:38,720 --> 00:09:43,160 Speaker 2: us for some sort of promotional material, you know West 181 00:09:43,200 --> 00:09:45,400 Speaker 2: worst case scenario, it could end up in a jurisdiction 182 00:09:45,440 --> 00:09:48,319 Speaker 2: it has no privacy protection whatsoever. You know, it is 183 00:09:49,040 --> 00:09:50,920 Speaker 2: a great concent and I've always said it's a great 184 00:09:50,920 --> 00:09:54,120 Speaker 2: consent to give these organizations DNA material because you never 185 00:09:54,240 --> 00:09:56,559 Speaker 2: quite know what's going to happen to it. I think 186 00:09:56,600 --> 00:10:00,440 Speaker 2: the real difference I see with DNA compared perhaps some 187 00:10:00,480 --> 00:10:03,000 Speaker 2: other personal information that we might give is if you 188 00:10:03,080 --> 00:10:06,120 Speaker 2: give someone your credit card number or your driver's license 189 00:10:06,200 --> 00:10:09,200 Speaker 2: or whatever, it can be highly damaging for that material 190 00:10:09,360 --> 00:10:12,000 Speaker 2: to get into the wrong hands, whether it's through a 191 00:10:12,040 --> 00:10:15,560 Speaker 2: cybersecurity breach, privacy breach, or because it's sold off to 192 00:10:15,640 --> 00:10:17,560 Speaker 2: someone that you didn't expect it to be sold to. 193 00:10:17,760 --> 00:10:19,120 Speaker 2: But at the end of the day. You can change 194 00:10:19,120 --> 00:10:21,280 Speaker 2: your driver's license, you can change your credit card, you 195 00:10:21,320 --> 00:10:23,839 Speaker 2: can change your passwords and so on. You can't change 196 00:10:23,840 --> 00:10:36,880 Speaker 2: your DNA. 197 00:10:37,040 --> 00:10:41,400 Speaker 1: So essentially these records could just be sold to the 198 00:10:41,480 --> 00:10:45,120 Speaker 1: highest better. Is there an opportunity, do you think because 199 00:10:45,160 --> 00:10:47,440 Speaker 1: you say that they're encrypted, right, so if I sent 200 00:10:47,480 --> 00:10:50,080 Speaker 1: off my spit to one of these companies, it would 201 00:10:50,120 --> 00:10:55,000 Speaker 1: be logged somewhere presumably x y Z female. Is there 202 00:10:55,040 --> 00:10:58,720 Speaker 1: a concern that if these were sold on to somebody 203 00:10:58,840 --> 00:11:02,800 Speaker 1: else that they could decrypt their information. 204 00:11:03,360 --> 00:11:06,120 Speaker 2: If they're sold to someone else, they will be decrypted 205 00:11:06,280 --> 00:11:09,960 Speaker 2: because you can't use the information when it's an encrypted form. 206 00:11:10,000 --> 00:11:12,400 Speaker 2: They're not going to buy a blob of ones and 207 00:11:12,520 --> 00:11:14,720 Speaker 2: zeros that they can't even see what they are. They're 208 00:11:14,760 --> 00:11:17,160 Speaker 2: going to decrypt it. The decryption key would be sold 209 00:11:17,200 --> 00:11:20,200 Speaker 2: along with the information to who revise it, because otherwise 210 00:11:20,240 --> 00:11:22,800 Speaker 2: you can't interact with the data. That's not to say 211 00:11:22,840 --> 00:11:25,200 Speaker 2: that they wouldn't hold it in an encrypted form. But 212 00:11:25,520 --> 00:11:28,760 Speaker 2: encryption is not a binary. Yes it's encrypted, no it's 213 00:11:28,800 --> 00:11:34,400 Speaker 2: not encrypted. Encryption scales along stratum from not encrypted to 214 00:11:34,920 --> 00:11:38,560 Speaker 2: industrial strength. You know it's held in the Pentagon and 215 00:11:38,640 --> 00:11:43,720 Speaker 2: it's encrypted in two and fifty six encryption methods. Having 216 00:11:43,760 --> 00:11:47,560 Speaker 2: said that, technology and computing power is increasing at such 217 00:11:47,600 --> 00:11:50,960 Speaker 2: a rate that what we thought of as very secure 218 00:11:51,240 --> 00:11:54,560 Speaker 2: methods of encryption or of storage ten years ago would 219 00:11:54,559 --> 00:11:57,959 Speaker 2: now be looked at as opensleather. You know, I read 220 00:11:58,000 --> 00:12:02,080 Speaker 2: the other day that Chinese researchers managed to use quantum 221 00:12:02,120 --> 00:12:06,880 Speaker 2: computing to potentially decrypt the most strong encryption that is 222 00:12:06,960 --> 00:12:10,640 Speaker 2: generally used by sites at the moment two for six AES. 223 00:12:11,160 --> 00:12:14,000 Speaker 1: So instead of female x y z, it goes on 224 00:12:14,000 --> 00:12:16,200 Speaker 1: to the new owner, and then that could they could 225 00:12:16,280 --> 00:12:19,960 Speaker 1: decrypt it and be like, oh, this is Chelsea Daniels's DNA. 226 00:12:20,120 --> 00:12:22,439 Speaker 2: Yeah, I mean in theory, there's two things that they 227 00:12:22,640 --> 00:12:26,120 Speaker 2: do to try and protect people's privacy. One is to encrypt, 228 00:12:26,160 --> 00:12:29,360 Speaker 2: and as we've talked about, potentially it can be decrypted. Secondly, 229 00:12:29,400 --> 00:12:34,360 Speaker 2: they anonymize it. So your DNA sequence wouldn't necessarily be 230 00:12:34,440 --> 00:12:38,000 Speaker 2: held with a label that sees Chelsea Daniels. It would 231 00:12:38,040 --> 00:12:41,640 Speaker 2: just be held as subject one two five seven six. Now, 232 00:12:41,840 --> 00:12:45,680 Speaker 2: the other difficulty that we're now facing with anonymization is 233 00:12:45,760 --> 00:12:49,640 Speaker 2: the ability for very strong computing power in the very 234 00:12:49,720 --> 00:12:53,800 Speaker 2: good matching of information, in particular with artificial intelligence to 235 00:12:54,040 --> 00:12:57,720 Speaker 2: match disparate pieces of information together and suddenly come up 236 00:12:57,760 --> 00:13:00,840 Speaker 2: with ay, well, person X was having a red car, 237 00:13:01,280 --> 00:13:04,959 Speaker 2: they logged onto a computer in this place at this time, 238 00:13:05,200 --> 00:13:07,920 Speaker 2: and they bought something with their credit card. Matching up 239 00:13:07,960 --> 00:13:10,240 Speaker 2: those three pieces of information, a person X must be 240 00:13:10,400 --> 00:13:12,840 Speaker 2: that person who owns that red car. Now that's a 241 00:13:12,840 --> 00:13:15,599 Speaker 2: lot more difficult, obviously, with DNA sequences and so on. 242 00:13:15,760 --> 00:13:20,400 Speaker 2: But there's clearly a move towards using high powered computers 243 00:13:20,400 --> 00:13:24,560 Speaker 2: to denonymize people so you can be targeted. So information 244 00:13:24,640 --> 00:13:27,640 Speaker 2: is getting it's getting much more granular in terms of 245 00:13:27,679 --> 00:13:30,640 Speaker 2: the ability to pick out a particular piece of information 246 00:13:30,720 --> 00:13:33,120 Speaker 2: and then link it to reidentify the person. 247 00:13:33,600 --> 00:13:39,440 Speaker 1: What would someone want with my DNA that information? Is 248 00:13:39,480 --> 00:13:42,160 Speaker 1: it all nefarious or is it kind of like a 249 00:13:42,200 --> 00:13:47,040 Speaker 1: pharmaceutical company wanting a great batch of DNA information to 250 00:13:47,200 --> 00:13:49,600 Speaker 1: then come up with a cure of the cancer or something. 251 00:13:49,720 --> 00:13:54,080 Speaker 2: Obviously, having research to produce cures to cancer using massive 252 00:13:54,320 --> 00:13:57,360 Speaker 2: DNA data sets is a good thing, no question about it, 253 00:13:57,480 --> 00:13:59,840 Speaker 2: if it's done in controlled circumstances. But as I said, 254 00:14:00,120 --> 00:14:02,440 Speaker 2: what we don't know who might end up being the 255 00:14:02,480 --> 00:14:05,120 Speaker 2: purchaser of this information, and it may not be someone 256 00:14:05,160 --> 00:14:07,679 Speaker 2: who is quite as ethical as we would all want. 257 00:14:07,840 --> 00:14:10,240 Speaker 2: This is one of the issues always with technology is 258 00:14:10,280 --> 00:14:12,480 Speaker 2: that you don't quite know where it's all going to 259 00:14:12,600 --> 00:14:12,959 Speaker 2: end up. 260 00:14:13,640 --> 00:14:16,080 Speaker 1: Another thing is when you're sending your DNA off to 261 00:14:16,120 --> 00:14:19,520 Speaker 1: one of these companies, you're not only rescinding your rights, 262 00:14:19,600 --> 00:14:22,760 Speaker 1: but in some cases your family's rights as well. Right 263 00:14:22,800 --> 00:14:26,320 Speaker 1: like the Green River killer was probably not too impressed 264 00:14:26,320 --> 00:14:28,800 Speaker 1: that his cousin of a cousin sent in one of 265 00:14:28,800 --> 00:14:30,280 Speaker 1: his dinner because that's how he got caught. 266 00:14:30,680 --> 00:14:34,000 Speaker 2: Right, So you know, other people can be identified. And 267 00:14:34,080 --> 00:14:36,720 Speaker 2: I mean that's the same with any law enforcement issue, 268 00:14:36,760 --> 00:14:39,320 Speaker 2: where you're often faced with the argument when you say 269 00:14:39,400 --> 00:14:42,240 Speaker 2: privacy is very important, you often get a pushback from 270 00:14:42,320 --> 00:14:45,320 Speaker 2: law enforcement people saying, well, if you've got nothing to 271 00:14:45,400 --> 00:14:48,880 Speaker 2: hide while you're worried about well, privacy is a human right. 272 00:14:49,000 --> 00:14:52,000 Speaker 2: Our right to be let alone is a right that 273 00:14:52,360 --> 00:14:55,160 Speaker 2: is important to all of us. We don't walk down 274 00:14:55,200 --> 00:14:57,880 Speaker 2: the road without any clothes on, and that's because there 275 00:14:57,880 --> 00:15:00,200 Speaker 2: are some things that we like to keep private. You know, 276 00:15:00,280 --> 00:15:03,280 Speaker 2: some people are less private than others, but it's all 277 00:15:03,280 --> 00:15:07,000 Speaker 2: a matter of choice. There's also with any technology, there's 278 00:15:07,040 --> 00:15:11,200 Speaker 2: also always the ability for mistakes to be made, false positives, 279 00:15:11,400 --> 00:15:14,640 Speaker 2: false identifications and sign and so forth. So you can 280 00:15:14,680 --> 00:15:18,320 Speaker 2: imagine these data sets being very because it's massive amounts 281 00:15:18,320 --> 00:15:21,680 Speaker 2: of data, you can imagine artificial intelligence being run across it. Now, 282 00:15:21,800 --> 00:15:25,360 Speaker 2: artificial intelligence is great, it's doing fantastic things, but it 283 00:15:25,440 --> 00:15:29,240 Speaker 2: makes mistakes. It's all it's doing is statistically predicting what 284 00:15:29,280 --> 00:15:31,760 Speaker 2: you expect to hear from it under an l l M, 285 00:15:31,960 --> 00:15:35,840 Speaker 2: so allow of the large language model, it's responding in 286 00:15:35,880 --> 00:15:38,760 Speaker 2: the way that it thinks you want based on the 287 00:15:38,760 --> 00:15:41,200 Speaker 2: prompt that you put in. And we don't know how 288 00:15:41,280 --> 00:15:44,840 Speaker 2: the artificial intelligence algorithms might be applied to these sorts 289 00:15:44,880 --> 00:15:48,240 Speaker 2: of things. So, you know, to take my health insurance example, 290 00:15:48,440 --> 00:15:52,000 Speaker 2: we don't know whether the algorithm that might be applied 291 00:15:52,040 --> 00:15:55,320 Speaker 2: across a huge data set would actually deliver the right result. 292 00:15:55,560 --> 00:15:58,040 Speaker 2: It may be that they've missed that it's misinterpreted the 293 00:15:58,040 --> 00:16:01,520 Speaker 2: gene sequencing and therefore, instead of giving me a clean 294 00:16:01,520 --> 00:16:03,360 Speaker 2: bill of health, has said, well, no, we think that 295 00:16:03,680 --> 00:16:06,480 Speaker 2: the algorithm thinks that you might be more likely because 296 00:16:06,520 --> 00:16:09,280 Speaker 2: of your DNA to have heart disease. That may be 297 00:16:09,280 --> 00:16:12,400 Speaker 2: completely wrong. So there are all sorts of difficulties. That's 298 00:16:12,440 --> 00:16:14,320 Speaker 2: not a function of the DNA, It's just a function 299 00:16:14,400 --> 00:16:16,440 Speaker 2: of how it works in the way and watch these 300 00:16:16,480 --> 00:16:18,560 Speaker 2: sorts of data sets may well will be used in 301 00:16:18,600 --> 00:16:19,040 Speaker 2: the future. 302 00:16:19,280 --> 00:16:23,080 Speaker 1: And one last question. If someone from your family turned 303 00:16:23,120 --> 00:16:25,280 Speaker 1: around and said they wanted to find out who your 304 00:16:25,280 --> 00:16:28,720 Speaker 1: ancestors were and they were going to use one of 305 00:16:28,760 --> 00:16:32,960 Speaker 1: these sides, what would you say to them? Don't just 306 00:16:33,040 --> 00:16:35,120 Speaker 1: straight up, no explanation, just no. 307 00:16:35,880 --> 00:16:38,560 Speaker 2: I've always said that. I mean, it's a choice, right, 308 00:16:38,680 --> 00:16:41,680 Speaker 2: it's I think the main thing is for people to 309 00:16:41,800 --> 00:16:45,640 Speaker 2: understand the potential downsides of doing it and then to 310 00:16:45,680 --> 00:16:50,280 Speaker 2: balance that against well, is that potential downside worth it 311 00:16:50,600 --> 00:16:53,360 Speaker 2: for the buzz that I'm going to get out of 312 00:16:53,400 --> 00:16:57,960 Speaker 2: finding that I have sixty percent Scottish ancestry. I don't 313 00:16:58,000 --> 00:17:00,160 Speaker 2: think it is. There are perhaps other ways of doing 314 00:17:00,240 --> 00:17:02,960 Speaker 2: that then using a DNA. And as I've said, the 315 00:17:02,960 --> 00:17:06,159 Speaker 2: thing that always strikes me as different about DNA and 316 00:17:06,200 --> 00:17:08,040 Speaker 2: the giving of DNA to other people is that you 317 00:17:08,119 --> 00:17:10,359 Speaker 2: cannot change it. But the other issue, of course is 318 00:17:10,400 --> 00:17:13,119 Speaker 2: that with any databases that there are in the world, 319 00:17:13,280 --> 00:17:17,800 Speaker 2: they are all susceptible to privacy breach, cybersecurity attack and 320 00:17:18,080 --> 00:17:20,240 Speaker 2: any organization that says, oh, I do you don't need 321 00:17:20,280 --> 00:17:22,960 Speaker 2: to worry about that, we are fully protected and we 322 00:17:23,080 --> 00:17:26,960 Speaker 2: have the best security in the world is absolutely telling porkies, 323 00:17:27,359 --> 00:17:30,760 Speaker 2: because all of the organizations that have been hacked all 324 00:17:30,800 --> 00:17:33,200 Speaker 2: said that right at the start. And as I say, 325 00:17:33,280 --> 00:17:37,400 Speaker 2: with computing power, with artificial intelligence, the people who are 326 00:17:37,680 --> 00:17:40,880 Speaker 2: obviously at the forefront of using that are criminals today. 327 00:17:40,920 --> 00:17:43,960 Speaker 5: Growing questions after a first of its kind data breach 328 00:17:44,080 --> 00:17:46,919 Speaker 5: targeting Jenny, how do you cite twenty three and me 329 00:17:47,240 --> 00:17:50,560 Speaker 5: twenty three and me confirm profile information whilst taken that 330 00:17:50,600 --> 00:17:55,520 Speaker 5: includes user names, passwords, gender, photo relatives in common and 331 00:17:55,560 --> 00:17:58,080 Speaker 5: the percentage of DNA you share with them. 332 00:17:58,119 --> 00:18:01,920 Speaker 2: So we can expect and we're all experiencing much more 333 00:18:01,960 --> 00:18:09,600 Speaker 2: sophisticated attacks on people's systems from hackers and criminals. And 334 00:18:10,359 --> 00:18:14,360 Speaker 2: what better pot of gold than a huge database of DNA. 335 00:18:14,720 --> 00:18:21,520 Speaker 1: Thanks for joining us, Rick. That's it for this episode 336 00:18:21,640 --> 00:18:24,640 Speaker 1: of The Front Page. You can read more about today's 337 00:18:24,720 --> 00:18:28,639 Speaker 1: stories and extensive news coverage at enzed Herald dot co 338 00:18:28,960 --> 00:18:33,040 Speaker 1: dot nz. The Front Page is produced by Ethan Seales. 339 00:18:33,440 --> 00:18:38,439 Speaker 1: Dan Goodwin is the sound engineer. I'm Chelsea Daniels. Subscribe 340 00:18:38,440 --> 00:18:41,320 Speaker 1: to the front page on iHeartRadio or wherever you get 341 00:18:41,320 --> 00:18:45,520 Speaker 1: your podcasts, and tune in tomorrow for another look behind 342 00:18:45,640 --> 00:18:46,520 Speaker 1: the headlines.