Google DeepMind Introduces SynthID to Combat Deepfakes and AI-Generated Imagery

Google DeepMind has unveiled SynthID, a watermarking technology aimed at identifying AI-generated images. The tool embeds an “imperceptible” watermark into the pixels of images generated by AI, making them easily identifiable by specialized detection tools. The technology is initially available to Google Cloud customers using the Vertex AI platform and the Imagen image generator.

As the 2024 election season looms in the U.S. and the U.K., the issue of deepfakes and AI-generated content has gained heightened attention. Google DeepMind’s CEO, Demis Hassabis, emphasized the increasing importance of systems that can identify and detect AI-generated imagery. “Every time we talk about it and other systems, it’s, ‘What about the problem of deepfakes?’” Hassabis said in an interview with The Verge on August 29, 2023.

How SynthID Works

SynthID works by embedding a watermark directly into the pixels of an AI-generated image. According to Hassabis, this watermark is “

robust to various transformations — cropping, resizing, all of the things that you might do to try and get around normal, traditional, simple watermarks.

The watermark does not alter the quality or the experience of the image but makes it easily detectable by DeepMind’s tools.

Availability and Future Plans

Initially, SynthID is being rolled out to a limited number of Vertex AI customers using Imagen, one of DeepMind’s latest text-to-image models. The technology was developed in collaboration with Google Research and is part of Google Cloud’s broader strategy to offer tools for creating AI-generated images responsibly.

Thomas Kurian, Google Cloud’s CEO, noted that the Vertex AI platform is experiencing rapid growth, making it an opportune time to launch SynthID. “The models are getting more and more sophisticated, and we’ve had a huge, huge ramp in the number of people using the models,” Kurian said.

Broader Implications

While SynthID is not the first tool designed to combat deepfakes, it joins a growing list of initiatives from major tech companies like Meta, Microsoft, and Amazon. These companies are increasingly investing in content verification solutions to maintain digital integrity. However, SynthID is not a “silver bullet to the deepfake problem,” according to Hassabis. The technology is still in its beta phase, and Google plans to refine it based on real-world testing and user feedback.

Image source: Shutterstock


Tagged : / / / / / /

YouTuber’s Channel Hacked for XRP Scams

The rise of cryptocurrency has brought about a wave of new investors, eager to get in on the action and reap the rewards. Unfortunately, this has also created a new target for hackers and scammers looking to make a quick profit. One recent example of this is the hacking of the popular YouTube channel DidYouKnowGaming.

DidYouKnowGaming is a YouTuber with 2.4 million subscribers who creates content related to video game trivia and history. However, the channel was recently hacked by an anonymous bad actor, who used it to promote XRP cryptocurrency scams. The hacker changed the channel’s profile and cover images to Ripple’s logo, in an attempt to lend legitimacy to the scam.

Fortunately, YouTube was quick to intervene and prevent further damage. They prevented the XRP hackers from interacting with the channel’s subscribers and worked with DidYouKnowGaming to regain access to his channel. However, this incident is just one example of a growing trend of hackers targeting YouTube channels to promote scams.

One of the largest YouTube creators, Linus Tech Tips, also recently reported losing access to his channels. While the exploit used by the hackers to gain access to YouTube accounts remains a mystery, it is clear that the threat to crypto investors from such hacks is prominent.

The rise of deepfakes only adds to this threat. Deepfakes are fake impersonation videos generated by artificial intelligence (AI) tools, and they have become increasingly prevalent in recent years. Hackers often create deepfakes of celebrities and entrepreneurs to misguide crypto investors and trick them into investing in scams.

For example, hackers have created deepfakes of Tesla CEO Elon Musk in the past, causing confusion among investors who thought he was endorsing a particular cryptocurrency. Concerns about deepfakes escalated even further when Chinese tech giant Tencent launched a new deepfakes creation tool, allowing users to impersonate anyone for a fee.

Crypto investors across the world use YouTube to learn about and research the world of cryptocurrencies, blockchain, and Web3. However, as the number of hacks and scams on the platform increases, it is important for investors to remain vigilant and take steps to protect themselves.

While YouTube and other platforms are working to prevent hacks and scams, investors should be wary of any investment opportunities that seem too good to be true. They should also be careful about the information they consume on the platform and take the time to research any claims made in videos or comments.

In conclusion, the hacking of DidYouKnowGaming’s YouTube channel is just one example of a growing trend of hackers targeting YouTube creators to promote scams. The rise of deepfakes only adds to this threat, and investors should remain vigilant when researching and investing in cryptocurrencies.


Tagged : / / / / /

AI Image Service Bans Xi Jinping Pictures

Midjourney, an artificial intelligence (AI) service that generates images from natural language descriptions, has recently found itself in the midst of controversy. The company has banned users from creating pictures of Chinese President Xi Jinping, citing concerns about deepfakes. Deepfakes are manipulated digital representations produced by sophisticated machine-learning techniques that can be used to spread false information and propaganda.

While Midjourney’s ban was intended to prevent the proliferation of deepfakes on its platform, users have found a way to circumvent the ban by creating deepfakes of Xi using other methods. Midjourney has responded by disabling access to its free trial version. Although the platform still allows the creation of images featuring world leaders, the Chinese President is notably excluded. Any attempt to generate an image with his likeness or even mention his name in a prompt is strictly prohibited by Midjourney.

Critics, like Sarah McLaughlin, senior scholar at the Foundation for Individual Rights and Expression (FIRE), have argued that the ban constitutes a form of censorship, undermining the fundamental principles of free speech and expression. While Midjourney’s intentions may have been to prevent the spread of deepfakes, the ban has sparked debate about whether it is ever appropriate to limit access to technology in this way.

In messages exchanged on the chat service Discord last autumn, Midjourney’s founder and CEO, David Holz, revealed that the firm had received complaints from local users about “various topics in different countries,” prompting them to block numerous related words. However, according to chat logs examined by The Washington Post, Holz refrained from listing the prohibited terms to prevent unnecessary controversy. Holz also mentioned that the prohibited words were not solely connected to China. Nonetheless, he recognized that China was a particularly sensitive matter, as political humor might put Chinese users at risk.

In response to the criticism, Midjourney has attempted to explain its decision. In a statement, the company stated that its ban on Xi Jinping images was not motivated by financial gain. Instead, the company claims that it was “for the greater good” to ensure that access to this technology was available to Chinese individuals. However, critics argue that this reasoning is flawed, as it assumes that Chinese individuals are incapable of using the technology responsibly.

Despite the ban, users of Midjourney’s platform have found a workaround by creating deepfakes of Xi using other methods. Some users have used the /imagine function and provided the full URL of an existing photo of Xi in the prompt. Others have used the /blend function to combine two existing photos. While these methods are not as seamless as generating an image directly from a description, they still allow users to create images of Xi.

In conclusion, Midjourney’s ban on Xi Jinping images has sparked debate about the limits of technology and free speech. While the company may have had good intentions, its decision has been criticized as a form of censorship. As the debate continues, it remains to be seen whether Midjourney will reconsider its decision or if users will continue to find ways to create deepfakes of Xi.


Tagged : / / / / /
Bitcoin (BTC) $ 26,290.04 0.58%
Ethereum (ETH) $ 1,588.53 0.20%
Litecoin (LTC) $ 64.40 0.16%
Bitcoin Cash (BCH) $ 210.33 1.77%