Meta's Plan To Protect Users From AI - Led Misinformation | NDTV Profit - video Dailymotion

NDTV Profit

#Meta's Sunil Abraham talks about the measures the company is taking to safeguard its users from #AI - led misinformation.   Watch him in conversation with Rishabh Bhatnagar.   For the latest news and updates, visit: ndtvprofit.com

Transcript

00:00 We're here at the sixth global AI leadership AI for India

00:04 Event by Asucham. I'm joined here by Sunil Abraham. He's the public policy director at Meta

00:11 Thanks so much Sunil for taking time out

00:14 I'd like to ask you what Meta is doing to sort of protect its users from AI and also AI led

00:21 misinformation

00:24 So I'll speak

00:26 Briefly about the various things Meta does on its own platform

00:30 When it comes to the risks

00:33 Associated with

00:37 Generative AI content

00:40 to begin with

00:42 When it comes to our own models

00:44 we have taken a

00:47 commitment to introduce

00:49 watermarks on a voluntary basis

00:53 Some of these AI tools are not available in India today, but when they are made available, they will be made available with these watermarks

01:00 The second change that happened last year and we have begun enforcement of that policy

01:07 from this year onwards is a change to a disclosure requirement specifically for political ads and

01:15 Also for ads around social causes what we require is that if there is any

01:22 digital

01:24 manipulation of the

01:26 Content associated with those ads then the advertisers have to make a voluntary disclosure

01:31 finally

01:33 especially for

01:35 Deceptive generative AI content circulating on whatsapp. We have worked with MCA

01:43 Which is leading consortium of fact-checkers and

01:50 We have upgraded the whatsapp tip line and now the fact-checkers

01:56 Manning

01:58 That tip line are able to also assess whether a particular piece of reported content

02:04 Has been generated or is

02:08 manipulated in any way

02:10 Sunil that was about Meta. You're also an active part of the asso champ

02:17 What is the industry the tech industry in India, especially big tech in India doing to sort of protect its users from AI?

02:24 So not just in India, but also

02:27 globally

02:29 corporations such as Meta

02:31 use

02:33 industry associations trade bodies

02:35 self-regulatory organizations to voluntarily impose certain standards upon ourselves

02:43 The best example of this is from the munity security munich security conference

02:49 We have entered into a tech accord with a variety of other global

02:54 companies

02:57 Part of the accord is a commitment

02:59 to develop

03:02 open source and also proprietary detection

03:06 capabilities

03:08 to find if

03:10 Deceptive generative AI content is being used in an election context

03:15 So that is one example of industry

03:18 Self-regulation there are many other examples such as the work that is happening the partnership

03:26 for AI and also

03:29 standards emerging metadata standards such as IPTC where meta also

03:36 Participates

03:38 (dramatic music)

03:41 (music)

Meta's Plan To Protect Users From AI - Led Misinformation | NDTV Profit

Category

Transcript

Recommended