Latest news with #Gen-AI
India Today
a day ago
- Entertainment
- India Today
Deepfakes are everywhere. But it's the harm, not the tech, that needs fixing
What do Elon Musk, Taylor Swift, Sachin Tendulkar, and Asaduddin Owaisi have in common? They have all been targeted by deep fake videos. Likelihood with celebrities across has been exploited for deep fakes promoting crypto schemes, financial scams, and fake you are reasonably active on social media, you might have already spotted AI-generated influencers that confuse you as to whether you are interacting with a real person or an AI bot. This phenomenon is so prevalent that we have started losing touch with what's is not believingGenerative AI (GenAI) has made it easy for anyone with an internet connection to create content that looks and sounds indistinguishably real. It is leading us to question our age-old wisdom that 'seeing is believing.' Gen-AI is altering our relationship with our visual and auditory senses. In the last century, we considered the camera the arbiter of truth and relied on this medium to tell us the we often forget that the same camera was also used in the last century to capture genuine moments of life, as well as for cinema and propaganda. Iconic images of Soviets capturing Berlin and the Toppling of Saddam Hussein's statue were all carefully staged images. Most social media content is also carefully planned, involving make-believe image seeing was not believing even before the Deep of SensesadvertisementIn fact, all our senses have been manipulated, and we have accepted this truth. We are constantly encouraged to laugh during Sitcoms using canned laughter, and most music, including vocals, has been heavily fabrics such as Nylon, Rayon, and PU leather have long hijacked our sense of touch. 99% of the vanilla we consume is not from orchid pods but from synthetic vanilla made from petrochemicals. Synthetic meat, Dalda, ground oil, etc., are other synthetic oils that have taken over our sense of fresh bread smell that you get from bakeries is not just from the freshly baked bread but also from maltol, diacetyl and yeast esters released by their ventilation the visual sense is not the only one being taken over by synthetic items. It is the latest one to be altered by do we regulate Deep Fakes?Whenever a new technology comes up, whether it was automobiles, the internet, or electricity, human civilisations have always grappled with this question. Deep fakes using Gen AI are no we then ban deep fakes altogether? Do we need to regulate tech or harms?A knife can be used in the kitchen, so it can also be used to harm a person. We do not regulate knives in the Kitchen, but have outlawed using them to have seen how our camera has been used in the past, both for good and bad. It has been used for spreading knowledge and information at the same time, for hatred and falsehood. With deep fakes, only the mode of origination has changed. Instead of a camera being used to commit harm, now an algorithm has been used to commit the same just like we didn't ban the camera for all it's good, we need not ban the Gen AI video or audio generation capabilities. We just need to continue regulating the harmful uses of this we concerned about an AI-generated adorable cat video? If his family consents, do we have any issues with a virtual replica of an erstwhile Bollywood actor acting in a new film?AI-generated videos need not be regulated altogether. We just need to regulate its bad uses, i.e., the generation of Gen AI content that harms a person or their instance, a deep fake that represents a celebrity in a scam or a pornographic video needs to be regulated, as should a fake video in which a politician makes a communally sensitive we need new laws for this?Interestingly, we may not need new laws to regulate deep fakes. Since we never regulated the camera but only its harmful uses, we can apply the same logic Bhartiya Nyaya Sanhita and Information Technology Act are equipped enough to deal with communally sensitive content, obscene content, child abuse materials, and defamatory can approach the police station if they are affected by deepfakes or other harms caused by Gen AI. We have seen cases where the police are unaware of how certain provisions under the Bhartiya Nyaya Sanhita or the Information Technology Act apply to Gen AI content as well. This emphasizes the need for education and greater TrustWhile we may not immediately enact new laws, there is indeed a need for transparency regarding synthetic visual and audio content (i.e., AI-generated content). Gen AI has arguably empowered both good and bad uses of the technology. The technology itself has amplified the problem of Deep Fakes as the tools are becoming increasingly government has already proposed that synthetic content should be obligatorily flagged to differentiate between genuine and AI-produced material. Many respected AI companies have adopted responsible AI principles, incorporating transparency into their outputs. These companies use a watermark in their outputs, helping users to determine whether the content is generated by AI. However, this practice is not universally adopted, leading to confusion among the public about whether content is AI-generated or real. The effectiveness of voluntary labelling remains limited, especially since malicious actors are unlikely to make progress, we need systems that automatically watermark or label AI-generated content at the point of creation—ideally at the model or platform level. Such watermarking mechanisms must be tamper-resistant and interoperable across platforms. Simultaneously, the government is rightly advocating for technical capabilities that enable tracing the originators of synthetic content. This could include embedded metadata, secure logs, or authenticated signatures that help identify the device, model, or user that produced the deep such measures become mandatory and widely adopted, enforcement agencies will be better equipped to trace and act against those misusing Gen AI tools to create deep as the camera was once feared for capturing the soul and invading privacy, GenAI is often perceived as a threat to our senses and reality. But over time, society learnt to distinguish the tool from its misuse. We regulated invasions of privacy and national security, not the act of photography itself. A similar approach can help us manage deep fakes by not banning the tool but addressing the harms arising from deep fakes.(Nikhil Narendran is a Partner in Trilegal's Bengaluru office and part of the TMT practice of the firm. He is a subject matter expert in the technology, media, and telecom communication space)- Ends(Views expressed in this opinion piece are those of the author)Trending Reel
Time of India
2 days ago
- Business
- Time of India
11 years of innovation: PubMatic's 2025 Hackathon
Innovation lies at the heart of PubMatic's culture. For over a decade, the company's internal hackathon has been a springboard for bold experimentation, collaborative energy, and transformative ideas. This year marked the 11th edition of the PubMatic Hackathon, held on May 6–7, and it proved to be the company's biggest and most impactful yet. Cross-functional collaboration at unprecedented scale This year's hackathon saw enthusiastic participation from across the organisation - from Engineering and Product to Business Operations and Client Services. With 100 teams and over 420 participants, the event stood out for both its scale and the diversity of perspectives involved. A key driver of success was the cross-functional collaboration. Teams combined design, development, product strategy, and business operations to craft solutions that were not only technically sound but also practical, scalable, and ready for real-world deployment. Over an intensive 48 hours, teams ideated, built, and presented their prototypes to a panel of judges. The quality, polish, and creativity on display set a new benchmark, making the judging process more challenging than ever. Spotlight on three strategic themes This year's hackathon centred on three strategic areas shaping the future of the industry: 1. Generative AI: Transforming experiences The rising adoption of Generative AI (Gen-AI) was unmistakable. Over 56% of all projects harnessed Gen-AI technologies - a powerful indicator of how rapidly this capability is becoming central to innovation. To put the scale into perspective: in just seven days, teams transacted a staggering 331.3 million tokens, averaging 47 million tokens per day. From faster prototyping to more intelligent outputs, Gen-AI was a catalyst for accelerated innovation. Participants explored a wide range of tools and frameworks, including: Gen AI-powered IDEs that streamlined development workflows Models from OpenAI and Gemini, tailored for advanced reasoning Agentic frameworks for building fast, autonomous prototypes From intelligent recommendation engines and AI-driven analytics to automated support tools, many of the Gen-AI-enabled ideas were not just futuristic - they addressed real, immediate challenges. PubMatic continues to embed Gen-AI across its ecosystem to create smarter platforms, automate workflows, and enhance customer experiences. 2. Monetisation: Unlocking new value streams Several teams focused on revenue optimisation - exploring ways to unlock greater value for customers and partners. Whether it was through refined bidding strategies or yield optimisation models, the goal was clear: drive measurable business outcomes through technology. 3. Observability: Building trust and transparency High-performing systems require deep visibility. This year, teams focused on elevating system observability to ensure greater transparency, reliability, and performance. From proactive monitoring to advanced alerting and diagnostics, the projects reflected a strong engineering culture centered on operational excellence. Looking ahead The 2025 Hackathon was a reminder of what drives PubMatic forward year after year: the pursuit of innovation through collaboration. It's not just about building great technology - it's about nurturing a culture that celebrates curiosity, experimentation, and continuous improvement. As the ideas from this year's hackathon begin to take shape in real-world implementations, PubMatic looks forward to its role in shaping the future of digital advertising. Head to PubMatic's Careers Page to check out open roles in engineering and beyond.
Business of Fashion
2 days ago
- Business
- Business of Fashion
How Investors Fell Back in Love With Fashion Tech
Alta was exactly what Amy Wu Martin was looking for. Last year, Wu Martin, an investor at Menlo Ventures, was searching for a startup that used AI to address the lack of personalisation in online shopping. After being introduced to Jenny Wang and trying out a beta version of her personal styling app Alta — which suggests looks built from their existing wardrobes, akin to the virtual closet in the 1995 film 'Clueless' — she wanted to help bring the app to the wider market. 'Every major e-commerce company probably has tried personalization to an extent, but doesn't really know you,' said Wu Martin, who previously invested in AI-driven companies like Gen-AI video platform Higgsfield and mental health startup Slingshot AI. 'Alta's thesis is they need to understand what's in your closet because that's the starting point.' It all led to Alta ultimately raising an $11 million seed round in June, led by Menlo Ventures, with participation from Aglaé Ventures, a venture firm backed by the Arnault family's holding company Financière Agache. Alta's fundraise is indicative of a wider trend among venture capitalists. The rapid adoption of AI is spawning a new class of startups that have encouraged Silicon Valley to give fashion tech — which they've sparsely invested in before — another look. In the past year, a number of startups that offer solutions for everything from design assistance to social shopping have collectively raised more than $100 million from serial tech investors like Thrive Capital, Menlo Ventures and Forerunner Ventures, among others. Luxury is also taking note of AI's impact on fashion: This year's LVMH Innovation Award winners, which go on to partner with the conglomerate's brands, included AI-driven content creation studio Omi and Kahoona, an AI-powered customer data platform. While 10 years ago, investors poured millions into e-commerce brands that pitched themselves as tech companies simply because they sold online, today they're attracted to those attempting to solve industry pain points via technology, such as virtual try-on and personalised shopping. But they're also the latest in a long line of businesses that have attempted to do the same and failed — it's been 30 years, and there's still no version of the 'Clueless' closet that's reached mass adoption. Still, venture capitalists are convinced that reality has never been closer. To get there, they're backing founders with rigorous tech chops, and unlike the DTC brands that were pressured to expand quickly, are more patient in helping new founders grow with gradual technological improvements. 'The nature of VC is that most of these companies will fail, and if anything has happened now … is that it's much easier to build a product,' Wu Martin said. 'The onus is really on the investor to really spend time with that founder; understand what the longer term defensible vision really is.' A New Investment Philosophy Investors today are less seduced by charismatic founders selling a dream — or fashion brands overselling their minimal tech capabilities. Instead, the companies securing funding have leaders with real expertise. Wang, for instance, is a software engineer who worked for companies like DoorDash. Search and discovery platform Daydream, which raised $50 million last June, was co-founded by Julie Bornstein, a former chief operating officer at online styling service Stitch Fix, while Cheryl Liu, a former technical product manager at Amazon founded Gen-AI design platform Raspberry AI, which raised $24 million in January. AI-powered virtual try-on app Doji, which raised $14 million in May, was created by former Google machine learning engineer Jim Winkens and Dorian Dargan, who helped build VR games at Meta. That experience allows founders to see both the big picture and the finer details of their businesses. Wang's technical skills made it easier for her to fine-tune Alta's software to improve the quality of its recommendations, according to Wu Martin; having a founder with those capabilities made her more confident investing in the platform. 'It was really important for us to find a team that was very AI native because we thought that actually solving the challenge of serving style is one of the hardest parts,' Wu Martin said. Investors are also keen on backing founders who have experience building similar platforms — particularly those that have already delivered for them. Bornstein, co-founder of AI-powered platform Daydream, which she describes as 'ChatGPT for fashion,' previously founded the personalised shopping startup, The Yes, which went on to be acquired by Pinterest in 2022. When the idea for Daydream first came to her, she reached out to Forerunner Ventures founder Kirsten Green, who had invested in The Yes. Green soon signed on as a partner, co-leading Daydream's seed round last June with Index Ventures, and participation from Google Ventures and True Ventures. 'When it comes to AI, you really want to bet on a founder,' said Frédérique Dame, a general partner at Google Ventures. 'When you find a founder that is as exceptional as Julie at understanding fashion intrinsically and how people shop, it's really incredible to be able to invest early.' Even fashion tech startups that don't use AI explicitly are getting a lift from the excitement around its potential to change the industry. Alexa von Tobel, founder of venture firm Inspired Capital, for example, said she invested in affiliate marketing software platform ShopMy in December 2023, in part because she predicts that a boom in AI-generated content will push consumers to seek referrals from humans — be it friends or their favourite influencers. Similarly, Try Your Best, a platform founded by Outdoor Voices founder Ty Haney that offers shoppers rewards from their favourite brands for sending referrals, posting about products and more, raised an $11 million Series A co-led by Strobe Ventures, which typically invests in Web3 companies, in June. 'The future of commerce will be more social, and there will be more real recommendations from real people, in a sea of a lot of noise,' Von Tobel said. A Slow Burn Founders that are trying to fundamentally change how people shop understand it will take longer to grow — and increasingly, so do investors. 'Now, companies are actually building a product with AI in mind in a way where you can push the agenda on personalisation and product discovery in a way you couldn't do before,' Dame said. 'It takes a while to build infrastructure that is flexible enough.' In the past year, Daydream reconfigured its system twice: it shifted from building its personalised search functions using OpenAI and receiving somewhat accurate results, to using multiple models, including Google's Gemini, to improve its search engine, Bornstein said. The company went into beta testing last October and officially launched its platform last month. 'It's a pretty complex system,' Bornstein said. 'To build it so it sort of works is actually quite easy; to build it so it really works is actually quite hard.' After launching, startups are introducing new features based on user demand in order to avoid building tools that won't resonate. Alta, for example, started generating packing lists for users' upcoming travel plans based on the city and climate after users asked for it. Later this year, it will begin recommending items a user needs to purchase ahead of a trip — another feature users requested, Wang said. It's not just making a product that works, but one that people will actually want to use. Last year, Alta was able to grow its user base to more than 10,000 after nabbing high-profile endorsements from investors like model Jasmine Tookes and television and podcast host Keltie Knight, but that is a miniscule number compared to the platforms people already go to for shopping and fashion inspiration: Instagram, TikTok and Substack. Getting more users to hop on board will likely be these apps' biggest challenges moving forward. 'Is it super fun and intuitive and exciting and easy to use? Are you thinking about it after you close out the app?' said Mary Korlin-Downs, who tracks the latest developments in the sector via her Instagram account, All Things Fashion Tech. 'Those are little factors that could totally set companies apart.'
Business Wire
6 days ago
- Business
- Business Wire
CommerceNext Announces 2025 Winners of the Annual CommerceNexty Awards
NEW YORK--(BUSINESS WIRE)--CommerceNext announced the winners of its 2025 CommerceNexty Awards on June 25 at the CommerceNext Growth Show in New York City. The annual awards program honors and celebrates the year's best and most innovative marketing and ecommerce programs. Recipients are selected through a combination of votes from a panel of industry leaders and an equally weighted popular vote from the community. "The caliber of submissions this year was extraordinary," said Veronika Sonsev, Co-Founder of CommerceNext. "What struck us most about this year's winners was their willingness to take creative risks. From AI-powered experiences to unexpected brand partnerships, they're showing the industry what's possible when you combine innovation with strategic thinking." Winners for 2025 were chosen across seven categories: Marketing Innovator of the Year M&M's won for its partnership with the Kate Spade New York team to launch the first-ever collaboration collection, Statement Candy. Best Use of Data or Technology in Marketing Coach was recognized for implementing interactive Gen-AI product previews online. Most Impactful Social Responsibility Campaign Patagonia won for its 'Unfashionable' campaign promoting durable, repairable and timeless products. Most Effective Brand Storytelling Bliss won for its 'Elder Millennial' campaign with comedian Iliza Shlesinger. Outstanding New Ecommerce Experience Paula's Choice was awarded for launching an AI-powered Skin Analyzer. Best Use of a New Channel Bath & Body Works was honored for tackling the challenge of selling fragrances online with TikTok Live Shopping. Women in Retail's Top Woman of the Year Award This award, in partnership with the Women in Retail Leadership Circle, recognizes a woman leader in retail who has contributed to the community and mentored others along her career path. This year's winner is Liana Thompson, Vice President of E-Commerce, Shutterfly. About CommerceNext CommerceNext is a community, event series and conference for marketing, ecommerce and technology leaders at retail and direct-to-consumer brands. In addition to the annual CommerceNext Growth Show, CommerceNext produces and hosts two executive retreats, regional networking events, webinars, industry research and an executive dinner series. CommerceNext is part of CloserStill Media, which owns the largest global ecommerce events portfolio, including Ecommerce Expo UK, E Show Madrid, E Show Barcelona and Ecommerce Expo Asia. To learn more, visit CommerceNext. About Women in Retail Leadership Circle/NAPCO Media Women in Retail Leadership Circle is comprised of successful, motivated and passionate women from the entire retail and brand ecosystem who have a strong belief that investing in themselves and in other women is good business. From CEOs with years of experience to directors working their way up, the common thread is women coming together to learn from each other and help each other succeed. Focused on both business strategy and executive professional development, WIRLC offers women leaders the opportunity to network, discuss industry issues, and learn how to become better leaders. The energized core of our community stems from our dedicated members who believe and support our mission. Beyond directly connecting with the industry's most elite women leaders, benefits of membership include discounted event registration, exclusive member-only content, a private networking forum, and so much more. Learn more at NAPCO Media, WIRLC's parent company, has been a leading information source in the markets it serves since 1958. NAPCO Media continues to evolve with information distribution methods and specializes in the creation and cross-channel distribution of exceptional content. NAPCO Media's mission is to build a community between its audience and its clients.
Business Standard
23-06-2025
- Business
- Business Standard
Coforge collaborates with Duke's Fuqua School of Business
To explore and accelerate adoption of Gen-AI in enterprises Coforge and Duke's Fuqua School of Business have come together to explore and expand the adoption of Generative AI in the enterprise space. Together, Coforge and Fuqua's business school students researched and identified powerful use cases of AI across multiple industries to drive transformative impact. This collaboration enabled the students to translate academic experience into industry applications, creating valuable business insights and opportunities. Students under the Fuqua Client Consulting Practicum (FCCP) were guided by both faculty members and business and technology experts from Coforge, specializing in areas such as Gen AI in Wealth Management, Gen AI in Travel, Gen and AI in Customer Service. The teams delivered strategic industry analyses and recommendations that offer clients and prospects a consultative framework for business transformation through generative AI. Coforge collaborates with universities, research labs worldwide on emerging technology areas such as AI, machine learning, and data; and provides students with opportunities to solve real world business cases. Vic Gupta, Executive Vice President, Coforge said, Duke University and the Fuqua School of Business has a very strong technology curriculum and attract the best minds from across the world for its management program. Collaborations with universities provide us with bright individuals who bring in creative thinking and fresh approach to industry challenges, guided by the faculty members and mentored by industry experts at Coforge, we collectively build real world solutions for our clients.
