- DG-DeepfakesIDV - mailman.kantarainitiative.org

AI DeepFake IDV meeting agenda
by Jay Meier 08 May '24

08 May '24

Hello all. I hope everybody is doing well. As we are all aware, we are painfully behind on our research production. Today, I would like to focus on troop-rallying, with hopes we can establish a target date for completion. Interesting thought….it may be that the reason we seem to have stalled on content creation is because we have conceptually reached our goal. In other words, it may be that there isn’t much else to contribute. The fact is that the subject is nascent and evolving. With that, we have a rough idea of content flow. We have lots of vocabulary, examples of attack vectors and a semblance of primary defense strategies. It may be that we simply are at the point of just rough drafting content. 1) So, lets first poll the group, for some feel of whether we are, in fact, at the end of the content contribution phase and whether we should start drafting (or whatever). 2) Next, assuming we agree that we are at that point, we need to zero in on how this content will be presented. Max has made several excellent points about avoiding traditional “white-paper” content. So, if not that, how will the content be presented? 3) Lastly, we need to start a conversation about what we want to be when we grow up. Assuming we do publish something official, what will the group become? Wednesday: Join The Zoom Meeting <https://us06web.zoom.us/j/86156446139?pwd=NVZJcVFOZmc3a1NWS0J6SURmc1VKUT09> Meeting ID: 861 5644 6139 Passcode: 808675 Meeting coordinates TIMES: 12:00PM Eastern Time on Wednesdays Jay Meier Senior Vice President - North American Operations jay(a)facetec.com | 612-978-3687 www.FaceTec.com <http://www.facetec.com/> | www.Liveness.com <http://www.liveness.com/> | www.SpoofBounty.com <http://www.spoofbounty.com/> Confidentiality Notice: This message is intended only for the use of the Addressee and may contain information that is privileged and confidential. If you are not the intended recipient, you are hereby notified that any dissemination of this communication is strictly prohibited. Please erase all copies of the message and its attachments.

1 0

Deepfakes content added to Source Material page
by mchaudoin＠mindspring.com 02 May '24

02 May '24

FYI – added some items to the resources page people might find interesting. * Webinar from American Banking Association: Deepfakes and Identity Fraud: Safeguarding Your Business in the Age of Generative AI. * Webinar PDF w/ contents from David Mattei, Datos Insights and Eric Lavine, Socure (link <https://kantara.atlassian.net/wiki/download/attachments/261128280/Deekfakes…> ) * Whitepaper from Socure: Unmasking document and biometric identity fraud. (link <https://kantara.atlassian.net/wiki/download/attachments/261128280/Socure%20…> ) * MIT Technology Review Article: AI Startup Synthesia w/ Hyperrealist Deepfake. link <https://www.technologyreview.com/2024/04/25/1091772/new-generative-ai-avata…> Regards, -Mike

1 0

Meeting Agenda addition
by Jay Meier 01 May '24

01 May '24

Hi again folks. We do have one thing to discuss, during todays meeting. NIST has issued an RFI, regarding The Administrations efforts to understand and regulate AI. Given where Kantara lives, relative to NIST, and since we are contemplating AI-born deepfakes, and potentially or certainly AI-born countermeasures within our group, we might have something to say about all this. We might consider responding to this RFI. https://content.govdelivery.com/accounts/USNIST/bulletins/399775f Thanks! Jay Jay Meier Senior Vice President - North American Operations jay(a)facetec.com | 612-978-3687 www.FaceTec.com <http://www.facetec.com/> | www.Liveness.com <http://www.liveness.com/> | www.SpoofBounty.com <http://www.spoofbounty.com/> Confidentiality Notice: This message is intended only for the use of the Addressee and may contain information that is privileged and confidential. If you are not the intended recipient, you are hereby notified that any dissemination of this communication is strictly prohibited. Please erase all copies of the message and its attachments.

2 1

DeepFake/IDV Group Meeting Agenda
by Jay Meier 01 May '24

01 May '24

Good morning! Hope everybody is well. We don’t have anything formally scheduled for todays meeting. Lets plan on circling up on content production. We are behind schedule and need to start cranking. Lets talk about using the AI content production tools to get up to speed. Wednesday: Join The Zoom Meeting <https://us06web.zoom.us/j/86156446139?pwd=NVZJcVFOZmc3a1NWS0J6SURmc1VKUT09> Meeting ID: 861 5644 6139 Passcode: 808675 Meeting coordinates TIMES: 12:00PM Eastern Time on Wednesdays Jay Meier Senior Vice President - North American Operations jay(a)facetec.com | 612-978-3687 www.FaceTec.com <http://www.facetec.com/> | www.Liveness.com <http://www.liveness.com/> | www.SpoofBounty.com <http://www.spoofbounty.com/> Confidentiality Notice: This message is intended only for the use of the Addressee and may contain information that is privileged and confidential. If you are not the intended recipient, you are hereby notified that any dissemination of this communication is strictly prohibited. Please erase all copies of the message and its attachments.

1 0

ChatGPT / Provenance Content
by Joey Pritikin 25 Apr '24

25 Apr '24

Hi everyone, As we discussed at this week's meeting, ChatGPT could be a useful tool for us to initiate content creation. While I'd never take its responses blindly, I do find that it is an effective research assistant and that I can significantly compress the time required to output quality content. Here is an example <https://docs.google.com/document/d/1yq3Of9kuYFq4j2iyIAJ_zKVrD1hnz-0O9WftM-x…> of using ChatGPT to create a framework for discussing the importance of and approaches to ensuring provenance that I literally did in <10 minutes. (Feel free to grab content from this if you like or drop comments in the doc - mostly I'm using this as an example of what we *could *do with the tool.) It would be great if we could get an official position from Kantara about whether ChatGPT or other generative AI-based content is acceptable as the basis for published works. I offer it as an approach, but look to leadership to provide guidance on its acceptability. Best, Joey -- Joey Pritikin Chief Product Officer +1 650 799 4611 joey(a)paravision.ai

1 0

DeepFake Weekly Meeting Agenda
by Jay Meier 24 Apr '24

24 Apr '24

Hello Everyone! I hope you are well. We skipped our meeting last week. It sounds like Denny may not be able to attend this week, but I intend to host the meeting. So here goes…. If you recall, during our weekly group meeting two weeks ago, we discussed alternative methods to combat AI-born deep fakes. We seemed to agree that we would characterize our workings in terms of “the Three P’s”; Policy, Protection & Provenance. During the call, Sir Bachenheimer referred me to Eric Scouten (Adobe), as a potential expert on data and data-object provenance. I agreed to visit with Eric, to discuss whether he should/could/would participate in the Kantara group. I visited with him and he gave me this link, which he believes summarizes Content Authenticity work, generally. https://ericscouten.dev/2024/content-authenticity-101/ According to Eric, content/data-object provenance regards ensuring the validity of such content/object, as well as a verifiable record of any potential edits to that content/object, as well as a verifiable record of how that content/object travels through the digital world. He insisted it really is not about identifying or defending against deep fakes at all. In fact, conversely, he suggested that Content Authenticity would rely on Digital Identity Credentialing, Verification and Authentication as part of the content authentication process. So, we agreed that I’d ask the group if we’d like to invite him to present to us, for our general education, but that he would not join or contribute to the group. There are some who disagree with him about this, including members of our group. So, today, during our Meeting, we need to discuss this and, hopefully, we can vote on whether to pursue Content Authentication or not, as a function of our work. Wednesday: Join The Zoom Meeting <https://us06web.zoom.us/j/86156446139?pwd=NVZJcVFOZmc3a1NWS0J6SURmc1VKUT09> Meeting ID: 861 5644 6139 Passcode: 808675 Meeting coordinates TIMES: 12:00PM Eastern Time on Wednesdays Jay Meier Senior Vice President - North American Operations jay(a)facetec.com | 612-978-3687 www.FaceTec.com <http://www.facetec.com/> | www.Liveness.com <http://www.liveness.com/> | www.SpoofBounty.com <http://www.spoofbounty.com/> Confidentiality Notice: This message is intended only for the use of the Addressee and may contain information that is privileged and confidential. If you are not the intended recipient, you are hereby notified that any dissemination of this communication is strictly prohibited. Please erase all copies of the message and its attachments.

1 0

Reinforcing the need for liveness detection
by Andrew Hughes 18 Apr '24

18 Apr '24

Here’s one you don’t see every day… Liveness detection for an in-person transaction. Apparently the bank teller was able to detect non-liveness of the customer! https://www.theguardian.com/world/2024/apr/17/brazil-woman-corpse-bank-fraud <https://www.theguardian.com/world/2024/apr/17/brazil-woman-corpse-bank-frau…> Andrew Hughes CISM m +1 250.888.9474 AndrewHughes3000(a)gmail.com https://www.linkedin.com/in/andrew-hughes-682058a

2 1

London UK - presentation attack hotspot?
by Andrew Hughes 16 Apr '24

16 Apr '24

In London for a few days and look what I found at a shop near Paddington… Pretty good looking presentation attack instruments if you ask me! And a real bargain! Andrew Hughes CISM m +1 250.888.9474 AndrewHughes3000(a)gmail.com https://www.linkedin.com/in/andrew-hughes-682058a

4 3

DeepFake Meeting Today
by Jay Meier 10 Apr '24

10 Apr '24

Hello All….I hope you are well. Both Denny and Andrew (I believe) will miss todays meeting. Denny, I and Andrew will miss next weeks meeting also. Despite that we don’t have much to discuss, on the Agenda, I will admin todays meeting. Next week, however, I suspect we will likely cancel the meeting. We don’t have a lot to discuss today. However, we can go through topic updates or continue with discussions about potential broadening of our group focus. Wednesday: Join The Zoom Meeting <https://us06web.zoom.us/j/86156446139?pwd=NVZJcVFOZmc3a1NWS0J6SURmc1VKUT09> Meeting ID: 861 5644 6139 Passcode: 808675 Meeting coordinates TIMES: 12:00PM Eastern Time on Wednesdays Add the group calendar to your personal calendar. <https://calendar.google.com/calendar/ical/c_fd984e780e396aa17f933c5bf938fb4…> (meeting is pinned to Eastern Time for Daylight Saving Time shifts - use the Time zone converter <https://www.timeanddate.com/worldclock/converter.html?iso=20230831T160000&p…> tool to convert to other time zones) Jay Meier Senior Vice President - North American Operations jay(a)facetec.com | 612-978-3687 www.FaceTec.com <http://www.facetec.com/> | www.Liveness.com <http://www.liveness.com/> | www.SpoofBounty.com <http://www.spoofbounty.com/> Confidentiality Notice: This message is intended only for the use of the Addressee and may contain information that is privileged and confidential. If you are not the intended recipient, you are hereby notified that any dissemination of this communication is strictly prohibited. Please erase all copies of the message and its attachments.

1 0

Re: How to detect AI deepfakes - The Washington Post
by maxine most 05 Apr '24

05 Apr '24

Free gift https://wapo.st/49q5tHb C. Maxine Most Principal, Acuity Market Intelligence +1 720 530 5836 640 W. Linden St, Louisville, CO 80027, USA www.acuitymi.com [signature_2563789463]<https://www.linkedin.com/in/maxinemost/> [A close up of a logo Description automatically generated] <https://twitter.com/cmaxmost> [signature_859388099] From: Andrew Hughes <andrewhughes3000(a)gmail.com> Date: Friday, April 5, 2024 at 11:43 AM To: maxine most <cmaxmost(a)acuity-mi.com>, Jay Meier <jay(a)facetec.com>, Prvu, Denny (He/Him/His) <denny.prvu(a)rbc.com> Subject: Fwd: [DG-DeepfakesIDV] How to detect AI deepfakes - The Washington Post Paywall for me... Are any names named? Should they know about our group? ———————— Andrew Hughes CISM m +1 250.888.9474 AndrewHughes3000(a)gmail.com<mailto:AndrewHughes3000@gmail.com> ---------- Forwarded message --------- From: maxine most <cmaxmost(a)acuity-mi.com<mailto:cmaxmost@acuity-mi.com>> Date: Fri, Apr 5, 2024 at 7:55 AM Subject: [DG-DeepfakesIDV] How to detect AI deepfakes - The Washington Post To: dg-deepfakesidv(a)kantarainitiative.org<mailto:dg-deepfakesidv@kantarainitiative.org> <dg-deepfakesidv(a)kantarainitiative.org<mailto:dg-deepfakesidv@kantarainitiative.org>> https://www.washingtonpost.com/technology/2024/04/05/ai-deepfakes-detection/ C. Maxine Most Principal, Acuity Market Intelligence www.acuitymi.com<http://www.acuitymi.com> +1 720 530 5836 _______________________________________________ A Community Group mailing list of KantaraInitiative.org DG-DeepfakesIDV mailing list -- dg-deepfakesidv(a)kantarainitiative.org<mailto:dg-deepfakesidv@kantarainitiative.org> To unsubscribe send an email to staff(a)kantarainitiative.org<mailto:staff@kantarainitiative.org> List archives -- https://mailman.kantarainitiative.org/hyperkitty/list/dg-deepfakesidv@kanta… ______ Group wiki -- https://kantara.atlassian.net/wiki/spaces/DFDG/overview

7 11