The Cookie Jar, July 2023
BoBWorld robs user data; Mega fines for bad ‘behaviour’; Big Tech’s Oppenheimer moment; Meta ‘Thread’-ening Twitter; Worldcoin’s ‘ick factor’; Llama 2 unleashed.
News Chunks
Banking on numbers
Bank of Baroda, India’s second-largest government-owned Bank, tampered with customers’ account details to boost the metrics of their online banking app, BoBWorld. Bank employees allegedly linked accounts to any mobile number they could gather to sign up for the app, then deregistered these users and used the same number to sign other accounts, compromising the security of several thousands of customers. The much anticipated Data Protection Bill, to be introduced in the Monsoon session of Parliament, allows laypersons to complain to the Data Protection Board of India if they believe their data has been compromised. If the Board finds a data breach, a fine of up to ₹ 500 crore can be levied.
Mega fines for bad ‘behaviour’
After attracting a $5 billion privacy and data security penalty from the FTC, and a €1.2 billion GDPR fine from the European Data Protection Board, among others, Meta’s bad ‘behaviour’ continues. Datatilsynet, Norway's data protection authority, has temporarily banned Facebook and Instagram from running behavioural advertising in Norway without informed user consent, or else face fines of up to 1 million kr per day.
Meanwhile, US authors are suing OpenAI and Google for copyright infringement as multiple claims of scraping original creative works to train proprietary AI models without permission surface. Source-code owners and visual artists are also suing these companies over certain codes and images that have been used to train these models.
Big Tech’s ‘Oppenheimer’ moment
The White House announced that seven top tech companies including Amazon, Google, Meta, Microsoft, Anthropic and OpenAI have committed to taking concrete safeguards to develop “safe, secure and trustworthy” technologies. Recognising the immense pace of new tech development, the companies have committed to 8 measures including watermarking, red-teaming, external and independent oversight and research among others. While red-teaming and security tests are existing practices, the US Federal Trade Commission will have its hands full getting the companies to deliver specific outcomes, especially on privacy and copyright, on a timeline, given these are yet to be drawn up.
How does it affect us in India? The same way that an arms race for nuclear power affects all countries. No country would want to be at the receiving end of the potential harms that a potent technology developed and deployed by a few, can unleash. Hear Christopher Nolan amplify AI Researcher’s concerns about Silicon Valley’s ‘Oppenheimer’ moment.
‘Thread’ with caution
Threads - Meta’s answer to a crumbling Twitter gained much attention very quickly. Even as Elon Musk and Mark Zuckerberg make nasty exchanges raking up engagement for their respective platforms, Instagram head Adam Mosseri stated Threads isn’t specifically trying to replace Twitter, and that although it is inevitable for news and politics to show up on the platform, the company isn’t going to actively encourage it. Bad actors have already taken to threads and Threads’ policies on hate, harassment and extremism remain obscure. ChatGPT, the not-so-old kid on the block to peak in record time has in the meantime, hit its first slump, and a Stanford study also established a steep decline in its performance.
“A clear ick factor…”
…is how Sam Altman himself described the identity verification process of his latest blockchain-traceable Worldcoin digital token. It would sound dystopian to many that Worldcoin will use a physical orb to scan people’s irises to distinguish humans from robots due to the surge of AI. With a reported $100 billion funding and a staggering $3 billion valuation, Worldcoin is on its way to scan about 1 billion people’s eyes in 2023. You heard that right. On one hand, the AI arms race leads to reckless deployment of potent technologies into social and economic ecosystems, and then the same founders come up with dystopian solutions that will further erode privacy and data security.
Those who love to ride a fad, (yes, web3 and crypto fans included) are sure to be early adopters and deliver on the project that Taigo Sada, Head of Product at Worldcoin, says “will make money”. However, with the UBI narrative hitched to this verification infrastructure, will those who tread more carefully be forced to give in as these walled-gardens close in?
No prob-Llama
Meta teamed up with Microsoft to launch their next-gen LLM - Llama 2. Llama2 is free for research and commercial purposes, and the model weights and starting code for the pretrained model are made open-source with the aim of making it ‘safer and better’. This also means bad actors can get their hands on this powerful tool. Llama 2 has been released with responsible use measures like responsible use guidelines and acceptable use policy in addition to existing responsible tech practices of red-teaming, and model cards. Meanwhile, Microsoft plans to charge a premium ($30 USD) for the generative AI features in its products, which will increase the monthly cost of the business grade version of Microsoft 365, by between 53% and 83%.
Find my phone
If you’ve ever panicked when you couldn’t find your phone - you’re definitely not the only one! The Department of Telecommunications, Government of India, has launched a new website called Central Equipment Identity Register (CEIR) where you can lodge a complaint about a lost electronic gadget with an International Mobile Equipment Identity (IMEI) at the local police station. You can also block a lost phone or unblock a recovered phone across all 37 states and union territories. To make things easier, CEIR has launched an app ‘Know Your Mobile” (KYM) which allows you to track the status of your phone.
Defib the apathy
The ongoing Manipur violence finally shook the nation with the emergence of videos of women bearing the brutal consequences of blatant, pre-meditated propaganda. India’s oft-favoured policy of internet ban as a method to curb anti-social and incendiary activities has often proven to be counter-productive with people’s access to fundamental needs and information choked, and disinformation exacerbated. This further stresses the need to tackle the ever-growing menace of information injustices of various kinds including freedom of speech and disinformation in a multi-ethnic country like India.
CDF Chunks
Litt’s new avatar!
Litt, CDF’s online information literacy initiative has a new look! Supported by FactShala, a media literacy network run by DataLEADS with support from Google News Initiative, the new Litt platform brings renewed commitment to empowering minds through information literacy – a critical skill for life and work in the 21st century.
Litt offers free, easy-to-attend courses that help you maximise the benefits and limit the harm of digital technologies. What’s more, Litt is also introducing courses in key Indian languages, and is building a library of information literacy and online safety resources from across India, for people from any State to know and access.
Individuals, educational institutions, NGOs, enterprises, PTA, or any other community can sign up or collaborate with us in taking these courses to more people in their region. Get in touch with us to make technology safe and productive for all.
CDF heads to Agami Makerfest
The OpenNyAI Maker Residency - a transformative and immersive five-day maker space - is underway at and in partnership with TinkerSpace, Kochi, with a goal of addressing critical challenges in the realm of law and justice through groundbreaking AI technologies, innovation and collaboration. Participants have the opportunity to delve into their ideas, prototype them, and receive guidance throughout the Residency.
Citizen Digital Foundation is delivering a sidebar on ‘Responsible tech philosophy and ethics practise in information technology’ at this event.
Bring your superpowers!
CDF is a non-profit, tackling techno-social issues like misinformation, online harassment, online child sexual abuse, polarisation, data & privacy breaches, cyber fraud, hypertargeting, behaviour manipulation, AI Bias, election engineering etc. We do this through knowledge solutions, responsible tech advocacy, and good tech collaborations under our goals of Awareness, Accountability and Action.