Connect with us

News

AI Compliance in Asia: Are Tech Giants Ready for the EU’s AI Act?

Explore the state of AI compliance in Asia as tech giants prepare for the EU AI Act, focusing on challenges and solutions in cybersecurity and bias.

Published

on

AI compliance

TL;DR:

  • Some prominent AI models struggle with EU regulations, particularly in cybersecurity and bias.
  • The EU AI Act introduces fines up to €35 million or 7% of global turnover for non-compliance.
  • LatticeFlow’s LLM Checker tool helps identify compliance gaps in AI models.

Artificial Intelligence (AI) is growing rapidly in Asia, with tech giants investing heavily in this transformative technology. However, as AI advances, so does the need for regulation. The European Union’s AI Act is set to shake things up, but are Asia’s tech giants ready? Let’s dive into the latest findings from LatticeFlow’s LLM Checker and explore the state of AI compliance in Asia.

The EU AI Act: A Game Changer

The EU AI Act is a comprehensive set of rules aimed at addressing the risks and challenges posed by AI. With the rise of general-purpose AI models like ChatGPT, the EU has accelerated its efforts to enforce these regulations. The AI Act covers various aspects, from cybersecurity to discriminatory output, and non-compliance can result in hefty fines.

LatticeFlow’s LLM Checker: Putting AI Models to the Test

Swiss startup LatticeFlow, in collaboration with researchers from ETH Zurich and INSAIT, has developed the LLM Checker. This tool evaluates AI models based on the EU AI Act’s criteria. The checker scored models from companies like Alibaba, Anthropic, OpenAI, Meta, and Mistral. While many models performed well overall, there were notable shortcomings in specific areas.

Discriminatory Output: A Persistent Challenge

One of the key areas where AI models struggled was discriminatory output. Reflecting human biases around gender, race, and other factors, this issue highlights the need for more inclusive and fair AI development.

  • OpenAI’s GPT-3.5 Turbo scored 0.46.
  • Alibaba Cloud’s Qwen1.5 72B Chat model scored 0.37.

Cybersecurity: The Battle Against Prompt Hijacking

Prompt hijacking is a type of cyberattack where hackers disguise malicious prompts as legitimate to extract sensitive information. This area also posed challenges for some models.

  • Meta’s Llama 2 13B Chat model scored 0.42.
  • Mistral’s 8x7B Instruct model scored 0.38.

Top Performer: Anthropic’s Claude 3 Opus

Among the models tested, Anthropic’s Claude 3 Opus stood out with the highest average score of 0.89. This model’s performance indicates that achieving high compliance with the EU AI Act is possible.

The Road to Compliance

Petar Tsankov, CEO and co-founder of LatticeFlow, sees the test results as a positive step. He believes that with a greater focus on optimising for compliance, companies can be well-prepared to meet regulatory requirements.

Advertisement

“The EU is still working out all the compliance benchmarks, but we can already see some gaps in the models. With a greater focus on optimising for compliance, we believe model providers can be well-prepared to meet regulatory requirements.” – Petar Tsankov, CEO, LatticeFlow

The Future of AI Regulation in Asia

As the EU AI Act comes into effect, Asian tech giants must prioritise compliance. Tools like LatticeFlow’s LLM Checker can help identify areas for improvement and guide companies towards developing more responsible AI models.

Comment and Share:

What steps is your organisation taking to ensure AI compliance with regulations like the EU AI Act? Don’t forget to subscribe. Share your insights and let’s discuss the future of AI regulation in Asia!

Author


Discover more from AIinASIA

Subscribe to get the latest posts sent to your email.

Life

AI at the Polls: Is Technology Steering the 2024 US Election?

As Americans cast their votes tomorrow, artificial intelligence will play a quiet but powerful role behind the scenes.

Published

on

AI at the Polls

TL;DR:

  • Campaign ads, social media feeds, and even “news” popping up in swing states are being shaped by AI’s invisible hand
  • Campaigns in 2024 aren’t just reaching voters; they’re diving deep into our digital footprints
  • AI brings campaigns closer to voters, it also makes it easier than ever to spread misinformation

A New Political Battleground—Inside the AI-Powered Election

As Americans cast their votes tomorrow, artificial intelligence will play a quiet but powerful role behind the scenes. Campaign ads, social media feeds, and even “news” popping up in swing states are being shaped by AI’s invisible hand. This isn’t just the next step in election tech; it’s a dramatic leap that could change the game forever. Is AI enhancing democracy, or are we giving it the keys to the whole democratic car?

1. Supercharging Campaigns: Microtargeting to the Extreme

Let’s face it—if you feel like your social media feeds are eerily personal, that’s not a coincidence. Campaigns in 2024 aren’t just reaching voters; they’re diving deep into our digital footprints to send messages so tailored they feel like personal letters. Thanks to AI, campaigns can slice the electorate into precise segments, tapping into anxieties, interests, and even specific local issues.

In battleground states like Arizona and Pennsylvania, this tech-driven targeting reaches a fever pitch. AI sifts through oceans of data—social media interactions, browsing habits, even purchase history—to craft ads that connect directly with you, personally.

“Campaigns are increasingly leveraging sophisticated machine learning algorithms to analyse vast quantities of voter data, refining their strategies with pinpoint accuracy,” notes MIT Technology Review (source).

With AI knowing so much, it raises an interesting (if slightly chilling) question: where’s the line between effective campaigning and outright manipulation?

Advertisement

2. The Double-Edged Sword: AI, Deepfakes, and Digital Misinformation

Here’s the darker side. While AI brings campaigns closer to voters, it also makes it easier than ever to spread misinformation. AI-generated deepfakes—fake videos that look so real you wouldn’t know they’re fake—have added a surreal twist to this election. Imagine seeing a video of a candidate saying something outrageous… and then realising it never actually happened.

“Deepfakes have made the spread of disinformation much easier and more convincing, raising concerns about the future of truth in politics,” the Brookings Institution warns (source).

AI’s power to create convincing fakes isn’t just a technical marvel; it’s a fundamental threat to truth in politics. Without strict regulations or ways to fact-check in real-time, we’re left wondering how many people will cast their vote based on a lie.

3. Predictive Polling: AI, Sentiment Analysis, and the All-Seeing Eye

If you thought AI was only influencing what you see online, think again. Polling has evolved far beyond traditional methods. This election, campaigns are using AI-driven sentiment analysis to tap into public moods in real time, keeping a pulse on issues that resonate with voters minute by minute.

“Sentiment analysis enables campaigns to see beyond traditional polling, observing shifts in public mood and identifying emerging concerns as they happen,” reports the Pew Research Center (source).

Advertisement

Let’s say economic concerns are heating up in Georgia; Trump’s team could amplify ads focusing on job growth in just hours. Or Harris’s camp could hone in on climate change in Michigan based on AI-driven insights from yesterday’s online conversations. This real-time fine-tuning isn’t just impressive—it’s a little mind-bending. Can polls really capture the pulse of the nation, or are we just seeing what AI’s algorithms want us to?

4. Mobilising the Masses: AI Nudges and Digital Persuasion

Getting people to the polls has always been crucial, and AI’s here to make sure more people than ever get nudged, reminded, and maybe even guilt-tripped into voting. AI-driven models predict not only who’s likely to vote but also who might need a little extra encouragement. Campaigns can then send targeted texts, emails, or even pop up on your social feed reminding you to “make your voice heard.”

The Atlantic remarks on AI’s power in mobilisation, stating, “AI has transformed voter outreach into an exact science, enabling campaigns to efficiently target and mobilise segments of the electorate that might otherwise stay home” (source).

For instance, Harris’s campaign has deployed AI to boost turnout among younger voters in key states, while Trump’s team uses it to rally dedicated supporters in traditionally red zones. AI doesn’t just follow you online; it’s practically waiting outside your door with a “Don’t forget to vote” sign. This kind of outreach raises a fascinating question about voter autonomy—are we freely deciding to vote, or are we being nudged by an algorithm?

5. Navigating the Ethical Minefield: Can Democracy Keep Up?

Here’s where it all gets tricky. While AI offers stunning capabilities for reaching, engaging, and mobilising voters, it also opens up new doors for potential misuse. From deepfakes to ultra-targeted political ads, AI is testing the limits of what’s fair game in political campaigns.

Advertisement

With regulations still trying to catch up, we’re left with a significant blind spot.

“Current frameworks for AI regulation are woefully inadequate, leaving a critical gap in safeguarding electoral processes,” states the Harvard Political Review (source).

AI has handed campaigns a powerful toolkit, but with great power comes… well, you know the rest. Without real oversight, there’s a real risk of crossing ethical lines, leaving voters questioning whether their choices are truly their own or just the echoes of an algorithm.

A Glimpse into Asia’s Future?

As AI’s influence in US elections becomes clear, Asia’s political landscape might not be far behind. In a region where social media is booming and governments increasingly leverage AI for everything from citizen services to surveillance, the potential for AI-driven election strategies is immense. Imagine a world where voter preferences in Tokyo, Jakarta, or Delhi are meticulously profiled, and campaign ads are hyper-personalised to every demographic, language, and cultural nuance. But here’s the question for Asia: with AI’s rapid adoption and limited oversight, who will control this powerful tool—governments, political parties, or the people? The US election offers a glimpse of how AI can shape democracy, but will Asia be able to harness this power responsibly, or could it open doors to unprecedented political manipulation? The stakes are high, and the path ahead remains uncharted.

Join the Conversation

How do you think AI will impact elections in Asia? Will it drive democracy forward or lead to new challenges in political manipulation? Leave a comment or subscribe for AI in Asia updates.

Advertisement

You may also like:

Author


Discover more from AIinASIA

Subscribe to get the latest posts sent to your email.

Continue Reading

News

Japan’s Nuclear Revival: Powering AI and a Greener Future

Japan is reviving nuclear power to fuel AI technologies and meet climate goals, with a focus on safety and public concern.

Published

on

Japan nuclear power AI

TL;DR:

  • Japan is restarting nuclear power plants to reduce coal dependency and meet climate goals.
  • The Kashiwazaki-Kariwa plant has new safety measures, including a 15-meter tsunami wall.
  • Nuclear power is essential for Japan’s plans to expand AI data centers.
  • Japan aims to increase nuclear power to 20-22% of its electricity mix by 2030.

The Nuclear Comeback

Japan is making a significant shift back to nuclear power. After the 2011 Fukushima disaster, the country shut down all its nuclear reactors. However, with climate change concerns and the need to power energy-hungry AI data centers, Japan is now reconsidering its stance on nuclear energy.

Safety First

At the heart of this revival is the Kashiwazaki-Kariwa nuclear plant in Niigata Prefecture. The plant has undergone extensive upgrades to prevent another Fukushima-like disaster. These include a 15-meter tsunami wall, backup power supply vehicles, and a new vent to filter out radioactive particles.

Masaki Daito, the plant’s deputy superintendent, believes these measures can prevent a similar accident. He told AFP, “We believe that a similar accident to Fukushima could be largely avoided. Japan now has the strictest (regulatory) standards in the world.”

The Energy Dilemma

Before the Fukushima disaster, nuclear power generated about a third of Japan’s electricity. Post-Fukushima, Japan increased imports of natural gas, coal, and oil to keep the lights on. However, this has not only been expensive but has also hindered Japan’s decarbonization efforts.

The E3G think-tank ranks Japan last among G7 nations in decarbonizing their power systems. Hanna Hakko, an energy expert at E3G, believes Japan could aim for renewables to generate 70-80% of its power by 2035. However, even in this scenario, the remainder would need to be covered by gas and nuclear energy.

Advertisement

The AI Factor

Japan’s ambitions in AI are another driving force behind the nuclear revival. AI data centers require vast amounts of energy, and nuclear power could provide a stable, low-emission source.

Prime Minister Shigeru Ishiba acknowledged this, stating, “Obviously, nuclear energy needs to be utilized.”

The Public’s Concern

Despite these plans, the memory of Fukushima still looms large. Many Japanese, like Mototsugu Oki, remain skeptical about nuclear power. “It is operated by human beings, and human beings naturally make mistakes,” he told AFP.

Moreover, making Fukushima fully safe is a monumental task. Engineers are still grappling with how to handle highly radioactive fuel and rubble.

The Path Ahead

Japan aims to increase nuclear power to 20-22% of its electricity mix by 2030. Nine reactors are currently online, with more set to join once they receive local approval.

Advertisement

However, the process is slow due to tougher safety rules. One restart was recently blocked because of earthquake risk.

Comment and Share:

What are your thoughts on Japan’s nuclear revival? Would you support nuclear power in your country to reduce carbon emissions and power AI technologies? Subscribe for updates on AI and AGI developments and share your thoughts in the comments below!

Author


Discover more from AIinASIA

Subscribe to get the latest posts sent to your email.

Continue Reading

News

Revolution Ahead: Microsoft’s AI Agents Set to Transform Asian Workplaces

Microsoft’s AI agents are set to transform Asian workplaces, boosting productivity and showcasing the practical benefits of AI.

Published

on

AI agents Asia

  • Microsoft is launching AI agents to handle tasks like client queries and supply chain management.
  • Copilot Studio will allow businesses to create AI agents without coding skills.
  • Early adopters like McKinsey and Clifford Chance are already using these agents.
  • AI agents could boost productivity and justify investments in AI technology.

Imagine a world where artificial intelligence handles your most tedious tasks at work. That world is just around the corner, as Microsoft prepares to launch autonomous AI agents next month. These virtual employees are set to revolutionise businesses, especially in Asia’s tech-savvy landscape. Let’s dive into what these AI agents can do, who’s already using them, and what it means for the future of work.

AI Agents: The Next Big Thing in Business Operations

Microsoft’s AI agents are designed to take on routine tasks, freeing up human employees for more creative and strategic work. Here’s what they can do:

  • Handle Client Queries: AI agents can manage customer service tasks, ensuring clients get quick and accurate responses.
  • Identify Sales Leads: These virtual employees can help sales teams by identifying and qualifying leads.
  • Manage Supply Chains: AI agents can streamline supply chain management, making it more efficient and reducing errors.

Copilot Studio: No Coding Skills Required

Microsoft’s Copilot Studio is a game-changer. It allows companies to create their own AI agents without needing any coding skills. Here’s how it works:

  • Pre-configured Agents: Microsoft will release 10 pre-configured AI agents designed for specific functions, like customer service and supply chain tasks.
  • Powered by Top AI Models: These agents will be powered by AI models developed by Microsoft and OpenAI, the creators of ChatGPT.

“Copilot Studio aims to simplify the creation of AI agents, requiring no technical knowledge from users.”

Early Adopters Leading the Way

Several companies are already using Microsoft’s AI agents to streamline their operations. Here are a few examples:

  • McKinsey: This consulting firm is using AI agents to schedule meetings and manage customer inquiries.
  • Clifford Chance: This law firm is leveraging AI agents to improve efficiency and reduce wasted time.
  • Pets at Home: This retailer is using AI agents to enhance workplace productivity.

Satya Nadella, Microsoft’s CEO, said, “These tools could revolutionize outsourcing by improving efficiency and reducing wasted time.”

The Future of AI Agents

Microsoft isn’t stopping at AI agents for basic tasks. They’re working on an agent capable of performing transactions on behalf of users. While there have been challenges, the technology is progressing rapidly.

Mustafa Suleyman, Microsoft’s head of AI, revealed that this transaction-capable agent could be available in a few months.

Addressing Concerns About Job Losses

There are concerns about the potential impact of AI agents on employment. However, Microsoft believes these agents will empower employees by taking over monotonous tasks.

Advertisement

Charles Lamanna, a corporate vice-president at Microsoft, said, “AI agents would enable employees to focus on more creative and meaningful work.”

The Potential Return on AI Investment

AI agents could provide a return on the significant investments tech companies have made in AI. While it’s uncertain how much revenue they will generate, they could showcase the practical benefits of AI.

Andrew Rogoyski, director at the Institute for People-Centred AI at the University of Surrey, noted that AI agents could help justify the billions of dollars poured into AI research and development.

What’s Next for AI in Asia?

Asia, with its tech-savvy population and rapid digital transformation, is poised to benefit greatly from Microsoft’s AI agents. Companies in the region could see increased productivity, improved customer service, and more efficient supply chains.

Comment and Share:

What tasks do you think AI agents could help with in your workplace? Share your thoughts below and don’t forget to subscribe for updates on AI and AGI developments. We’d love to hear about your experiences with AI technologies and your predictions for future trends!

Author


Discover more from AIinASIA

Subscribe to get the latest posts sent to your email.

Continue Reading

Trending

Discover more from AIinASIA

Subscribe now to keep reading and get access to the full archive.

Continue reading