DailyPost 2633
THE A.I. ACT
What fruitful results the 23-year-old IT Act brought to this country is for all of us see. Data Protection Act took its own sweet time, and social media remains the battlefield, with or without intermediary rules. While the technology onslaught has been the same across the globe, the congressional, legal and enforcement reaction has been tragically different in different countries. This single fact explains how the tech functions (meaning tech companies) in those geographies and what quantum it brings as revenue to the nation. While the European Union got into thinking of regulating Artificial Intelligence way back in 2018, rest of the world is still groping in the dark.
The first draft was released in 2021 and last Friday it was passed as the A.I. Act. With the passage of the A.I. Act has secured its place as the pioneers of tech legislation in the world. The technological breakthroughs post the first draft of the Act forced them to rewrite the law. ChatGPT made it imminent to put the law in place as early as possible. The initial version of the Act made no mention of the general purpose A.I. models like the ones that power ChatGPT. ChatGPT in a way has triggered a race of multimodal large language models, potent enough to rip apart our social, political and economic fabric.
EU has a track record of creating tech legislation, where is needed the most. It has created the gold standard of data protection law, the General Data Protection Regulation, GDPR. The European Parliament approved it on April 2016. It went into operation on May 25,2018. In the same spirit, this current agreement over the A.I. Act “solidifies one of the world’s first comprehensive attempts to limit the use of artificial intelligence.” Unregulated AI can turn out to be suicidal to say the least. This Act is bound to become the new global benchmark for different nation states to harness the potential benefits of this technology, while having all the safeguards in place.
Reckless automation of jobs, spreading disinformation and endangering national security can be some of the risks. The European policy makers have focused on AI’s riskiest uses by companies and governments. Large general purpose AI systems would need to have new transparency requirements. Generated by AI needs to be mentioned as a safeguard. Use of facial recognition would be restricted outside of safety and national security exemptions. There are strict penalty provisions as well. “Technological dominance precedes economic dominance and political dominance,” Jean-Noel Barrot, France’s digital minister said recently. India needs to put its act in order and EU’s A.I. Act can be a good reference law.
DOES AI FIND ANY MENTION IN OUR NATIONAL NARRATIVE? IF NOT, WE ARE HEADING TOWARDS BECOMING AN AI SLAVE NATION.
Sanjay Sahay
Have a nice evening.