AI continues to make headlines in the tech world, with new advancements and innovations being announced regularly.
Here’s a roundup of some of the latest news from the artificial intelligence world:
Poland’s ruling party – Law and Justice – is working on a draft of electronic communications law that would give authorities easier access to citizens’ emails and social media messages. The bill would require email service providers to make data available to the relevant court, including the IP address and content of emails, and would also make it easier for authorities to access conversations on messaging apps like WhatsApp and Messenger. The bill has been criticised by opposition politicians, who claim it would increase state surveillance. Experts also question the feasibility of the bill, given that companies based outside of Poland may not comply with requests for data.
For more information – click here.
A prestigious French university, Sciences Po, has banned the use of the AI chatbot – ChatGPT for student assignments, citing concerns about plagiarism. The university stated that anyone found to have used the chatbot would face severe sanctions, including expulsion from the school. The decision comes after it was revealed that ChatGPT had passed exams at a US law school. Other education facilities in other countries have also banned the use of ChatGPT and other AI tools in schools and universities, and some are changing exam formats to prevent cheating.
For more information – click here.
The U.S. Department of Commerce’s National Institute of Standards and Technology (NIST) has released its Artificial Intelligence Risk Management Framework (AI RMF 1.0), a guidance document for organizations using AI to manage AI-related risks. The AI RMF is divided into two parts: discussing AI risks and outlining trustworthy AI systems, and providing four functions to address AI risks. The framework is voluntary and is part of NIST’s larger effort to establish trust in AI. NIST will update the framework periodically with suggestions from the AI community and will launch a Trustworthy and Responsible AI Resource Center to help organizations put the AI RMF into practice. The agency released a companion AI RMF Playbook to suggest ways to use the framework.
For more information – click here.
The US congress member delivered a speech written by the online AI chatbot ChatGPT. It was the first time an AI-written speech was read in Congress. The speech was about a bill that would establish a joint US-Israel AI Center for AI research. The congress member hoped the speech would spur debate on AI and its challenges and opportunities.
For more information – click here.
Microsoft has created VALL-E, a text-to-speech AI model that can mimic someone’s voice from just a 3-second audio sample. It can match the speaker’s timbre, emotional tone and even the acoustics of the room. The model was trained on 60k hours of speech from 7k+ speakers. Results are mixed with some samples sounding realistic and others machine-like. Microsoft plans to improve the model by increasing its training data, but has not made the code open-source due to potential risks of misuse.
For more information – click here.
U.S. and EU announce AI agreement to enhance agriculture, healthcare, emergency response, climate forecasting and electric grid. The initiative will improve government operations and services, giving greater access to detailed AI models. The U.S. and EU data will remain separate but a common AI model will be built to improve results. The partnership is between White House and European Commission, other countries to join in coming months.
For more information – click here.
Author: Agata Konieczna
Credits:
Poland icons created by Anditii Creative – Flaticon
Ai icons created by pojok d – Flaticon
Framework icons created by Eucalyp – Flaticon
Government icons created by Freepik – Flaticon
