Edition#6

Hello Readers,

I trust you had a fantastic spring break. The burst of colors from cherry blossoms made mine truly delightful! Below, I'm sharing one of my favorite snapshots taken at the tidal basin in Washington, DC.

A lot is happening in our world! The General Assembly recently adopted a groundbreaking resolution on AI, led by the United States and supported by over 120 Member States. This resolution emphasizes the importance of AI systems complying with international human rights law and ensuring consistent protection of individuals' online rights, mirroring offline safeguards throughout the life cycles of AI systems.

In the US, Vice President Kamala Harris announced the completion of the 150-day actions directed by President Biden's AI Executive Order, marking a significant milestone in efforts to enhance AI safety, privacy, equity, innovation, and more.

Elon Musk made an intriguing prediction, suggesting that artificial general intelligence (AGI) could outsmart humans as soon as next year or by 2026. However, there's a catch: Musk highlighted a hurdle in the plan. Grok's version 2 model is facing delays due to a shortage of advanced chips, underscoring the crucial role of chip technology in AI progress. Nonetheless, he emphasized the critical importance of electricity availability in driving AI forward in the near future.

Furthermore, echoing similar concerns, Arm Holdings CEO Rene Haas issued a stark warning, highlighting the potential for AI data centers to consume a quarter of the U.S. power requirements by the decade's end if energy efficiency measures are not enhanced.

In other news, China has relaxed restrictions on cross-border data flow, aiming to provide clarity and guidance on data transfer requirements while balancing the need for data security with facilitating legitimate cross-border data flows.

Lastly, there are two intriguing developments in tech governance: a lawsuit filed against Apple in the U.S. District Court alleging unlawful maintenance and monopolistic practices in the smartphone market, and a proposed Assembly Bill in California aiming to grant employees the "right to disconnect" from work-related communications outside of their regular working hours.

Happy reading and stay tuned for more exciting updates in the next edition!

AI

General Assembly adopted landmark resolution on artificial intelligence

turkiye.un.org : The recently adopted resolution on artificial intelligence (AI) by the UN General Assembly underscores the importance of respecting, protecting, and promoting human rights in all aspects of AI development and deployment.

Led by the United States and supported by over 120 other Member States, the resolution emphasizes the need for AI systems to comply with international human rights law and mitigate risks to human rights. It asserts that the same rights individuals have offline must be safeguarded online, throughout the entire life cycle of AI systems.

Furthermore, the resolution recognizes AI's potential to contribute to achieving the UN's Sustainable Development Goals (SDGs) and calls for cooperation to bridge the digital divide between countries. Developing nations, in particular, face challenges in keeping pace with technological advancements, and the resolution urges support to enhance inclusive access and digital literacy.

The resolution is intended to complement existing UN initiatives and future efforts, including negotiations for a global digital compact and the work of the Secretary-General's high-level advisory body on AI. It reinforces the commitment to governing AI technology in alignment with shared priorities for sustainable development and equitable progress.

FACT SHEET: Vice President Harris Announces OMB Policy to Advance Governance, Innovation, and Risk Management in Federal Agencies’ Use of Artificial Intelligence:?

Whitehouse.gov : Today, Vice President Kamala Harris announced the completion of the 150-day actions tasked by President Biden's Executive Order on AI, marking a significant milestone in the administration's efforts to strengthen AI safety and security, protect privacy, advance equity and civil rights, promote innovation, and more. The White House Office of Management and Budget (OMB) issued its first government-wide policy aimed at mitigating AI risks and harnessing its benefits, fulfilling a core component of the Executive Order.

OMB's new policy directs federal agencies to address risks associated with AI use by implementing concrete safeguards to protect Americans' rights and safety. By December 1, 2024, agencies must implement measures to reliably assess, test, and monitor AI impacts, mitigate algorithmic discrimination risks, and provide transparency into AI usage. These safeguards will apply across various sectors, including health, education, employment, and housing.

?

Tesla's Musk predicts AI will be smarter than the smartest human next year

Reuters: Elon Musk, CEO of Tesla and co-founder of OpenAI, made a bold prediction on Monday, suggesting that artificial general intelligence (AGI) could surpass human intelligence as early as next year or by 2026. Musk emphasized that the development of AGI, which he defines as being smarter than the smartest human, is progressing rapidly.

Musk noted that the lack of advanced chips has hindered the training of Grok's version 2 model, highlighting the significance of chip technology in AI development. However, he also emphasized that the availability of electricity will be crucial in the coming years for AI advancement.

Switching gears to electric vehicles (EVs), Musk acknowledged the fierce competition from Chinese carmakers, describing them as the most competitive globally and posing significant challenges to Tesla. This recognition underscores the growing prominence of Chinese EV manufacturers in the global market and their potential impact on Tesla's dominance in the EV industry.

AI Could Consume 25% Of US Power By 2030: 'The More Information They Gather... The More Power It Takes,' Warns Arm CEO

Benzinga.com : Arm Holdings CEO Rene Haas issued a stark warning on Tuesday, highlighting the potential for artificial intelligence (AI) data centers to consume a quarter of the U.S. power requirements by the end of the decade without improved energy efficiency measures. To address this concern, Arm and its parent company, SoftBank Group, pledged $25 million towards a $110 million initiative aimed at supporting AI research at universities in the U.S. and Japan.

According to Haas, AI models such as OpenAI's ChatGPT already demand significant electricity, and this demand is expected to surge in the coming years. He emphasized that these models require an "insatiable" amount of power to operate effectively, with each interaction with ChatGPT consuming nearly ten times the electricity of an average Google search.

The International Energy Agency has also highlighted this issue, emphasizing the urgent need for more sustainable energy solutions to support the advancement of AI research. In response, Arm's funding will support a U.S.-Japan research collaboration involving Carnegie Mellon University and Keio University, focusing on innovative ways to reduce electricity consumption in AI technologies.

This initiative aligns with a broader strategy led by U.S. ambassador to Japan, Rahm Emanuel, who has orchestrated a $150 million U.S.-Japan quantum computing research program. Major tech companies like Amazon.com Inc. are also investing in this effort, supporting additional research endeavors in both Washington and Tsukuba

Google AI could soon use a person’s cough to diagnose disease:

Nature.com : A team led by Google scientists has developed a machine-learning tool called Health Acoustic Representations (HeAR) that shows promise in detecting and monitoring health conditions by analyzing sounds such as coughing and breathing. Trained on a massive dataset of millions of human audio clips, the AI system has the potential to diagnose diseases like COVID-19 and tuberculosis, as well as assess lung function.

What sets HeAR apart is its extensive training data and its ability to perform multiple tasks through fine-tuning.HeAR was adapted to detect COVID-19, tuberculosis, and characteristics such as smoking habits. Despite being trained on diverse sound sources, the model achieved promising results, outperforming existing models in some cases.?

The extensive data used by Google adds credibility to the research, according to experts like Ali Imran from the University of Oklahoma, who is developing an app named AI4COVID-19. His team plans to seek FDA approval for the app's diagnosis capabilities, as there are currently no FDA-approved tools for diagnosing diseases based on sound.

The field of health acoustics, also known as 'audiomics', holds immense promise for disease diagnosis, screening, and monitoring. Yael Bensoussan from the University of South Florida emphasizes the potential of voice as a non-invasive and low-resource biomarker for tracking health. She co-leads a research consortium focused on exploring voice as a biomarker for health monitoring.

While the Google-developed tool is still in its early stages, it represents a significant advancement in leveraging AI and machine learning for healthcare applications, potentially revolutionizing disease diagnosis and monitoring processes in the future.

Cross Border Data

China Eases Restrictions on Cross-Border Data Flows:?

Insideprivacy.com : After nearly six months since the initial draft was issued for public comments on September 28, 2023, several key changes and clarifications have been introduced in the Provisions:

1. Transfer of Non-Personal and "Non-Important" Data (Article 3): Cross-border transfers of data collected during activities such as international trade, academic cooperation, or cross-border manufacturing are exempted from transfer mechanism requirements if the data does not contain personal information or important data. This means that pre-transfer approval from the Cyberspace Administration of China (CAC) generally isn't necessary for such transfers.

2. Transfer of Important Data (Article 2): Data processing entities are required to identify and declare "important data" based on regulations to be issued by sectoral and/or local regulators. Companies are only required to undergo a security assessment for the transfer of "important data" if informed by the regulator or through public notice.

3. Exemption for Specific Categories of Transfer Purposes (Article 5): Transfers deemed necessary for contractual necessity, cross-border human resources management, or emergency situations are exempted from transfer mechanism requirements.

4. Exemption Based on "Negative Lists" in Free Trade Zones (Article 6): Local governments in free trade zones (FTZs) can propose negative lists for data subject to transfer mechanism requirements, similar to negative lists for foreign investment. Other types of data will generally be exempted from these requirements.

5. Exemption for Data Transit (Article 4): Data originating outside of China that merely transits through China without involving domestic personal information or important data is exempt from transfer mechanism requirements.

6. Thresholds for Adopting Transfer Mechanism (Article 7 and Article 8): Transfer of personal information by a critical information infrastructure (CII) operator or transfer of important data will trigger mandatory CAC approval, consistent with existing rules.

These Provisions aim to provide clarity and guidance on data transfer requirements in China, balancing the need for data security with facilitating cross-border data flows for legitimate purposes.

Privacy

US lawmakers unveil a plan to give all Americans a right to online privacy

CNN.com : Two prominent lawmakers in the United States have brokered a bipartisan agreement that could potentially grant all Americans a fundamental right to digital privacy. This deal, if successful, would establish a national law governing how companies collect, share, and utilize online data, akin to the European Union's General Data Protection Regulation (GDPR). The proposed legislation aims to address concerns regarding the unregulated space where Americans' personal data is often shared and sold without adequate oversight.

The agreement represents a significant breakthrough after years of stalled negotiations between Republicans and Democrats on this issue. However, it could potentially override stringent state-based privacy laws, such as those in California. The proposed legislation covers a wide range of entities, including data brokers, tech platforms, and telecom providers, but excludes small businesses and government contractors.

The lawmakers' proposed American Privacy Rights Act would prohibit the transfer of sensitive personal data to third parties without explicit user consent, except for specific purposes outlined in the bill, such as fraud prevention. It would also allow users to opt out of targeted advertising and require companies to collect only necessary data for their operations. Additionally, it would grant individuals the right to access, correct, or delete their data from a company's records.

In response to concerns about foreign adversaries accessing Americans' personal data, the legislation would require companies to disclose if data may be sent to or processed in countries like China or Russia. This issue has gained prominence amid worries about data security, particularly regarding platforms like TikTok and data brokers.

The agreement addresses longstanding disagreements between Republicans and Democrats over the scope of federal privacy laws, including whether they should preempt state laws and whether individuals should have the right to sue companies for privacy violations. The proposed legislation would preempt existing state privacy laws and enable individuals to sue companies for violating the law.

While hailed as a significant opportunity to establish a national data privacy standard, the legislation still faces hurdles, including approval from both lawmakers' committees and both chambers of Congress before reaching President Joe Biden's desk. Given the political landscape leading up to the 2024 election, passing such legislation may prove challenging.

Data from 73 million AT&T accounts leaked to dark web, company says

washingtonpost.com : Information from 73 million current and former AT&T accounts appears to have been leaked onto the dark web, the communications company said Saturday.

The data set was released about two weeks ago onto the dark web, which can be used for illegal activity by anonymous criminals. It may have included information such as Social Security numbers, phone numbers, email addresses, full names, dates of birth and mailing addresses, according to AT&T’s note to customers. The company said it reset the passcodes of 7.6 million current customers.

The company said it was notifying customers whose data was leaked and would pay for credit monitoring when applicable.

Many high-profile companies have suffered data breaches, including both of AT&T’s biggest rivals. Verizon has seen multiple reports of information leaks over the years, and T-Mobile has had at least eight incidents since 2018, including a 2021 instance in which hackers stole the data of millions of customers. This year, a hack at a third-party company may have exposed the account information of American Express cardholders.

Big Tech

Justice Department, states accuse Apple of holding a smartphone monopoly

justice.gov: In a recent development, a lawsuit has been filed against tech giant Apple in the U.S. District Court for the District of New Jersey. The complaint alleges that Apple has unlawfully maintained a monopoly in the smartphone market by imposing restrictive contractual terms on developers and withholding critical access points. It is claimed that these actions hinder competition and prevent the development of apps, products, and services that could offer consumers alternatives to the iPhone, promote interoperability, and lower costs.

The lawsuit, brought forth by the Justice Department and state Attorneys General, seeks to address concerns about Apple's alleged exploitation of its monopoly power to extract higher prices from consumers, developers, and other stakeholders. Attorney General Merrick B. Garland emphasized the importance of enforcing antitrust laws to protect consumers from unfair pricing and limited choices, asserting that no company, regardless of its size or prominence, is exempt from legal scrutiny.

Deputy Attorney General Lisa Monaco reiterated the government's commitment to upholding the rule of law, emphasizing that no corporation is above scrutiny. Acting Associate Attorney General Benjamin C. Mizer highlighted the significance of taking action against anticompetitive practices to safeguard economic justice and foster innovation.

Tech Governance

Bill proposed by Assembly member Matt Haney would give California workers the "right to disconnect"

cbsnews.com : Assembly Bill introduced by Assemblymember Matt Haney in California, proposes granting employees the "right to disconnect" from work-related communications outside of their regular working hours. The bill aims to protect workers from feeling compelled to answer emails or calls from supervisors during their off time. It requires both public and private employers to establish a workplace policy allowing employees to ignore such communications, with non-working hours determined by a written agreement. Violations could result in fines, with complaints handled by the California Labor Commissioner. While some concerns exist, Haney suggests that employers can address this by implementing clear policies on work hours and emergency communications. The bill aims to prevent burnout, and if passed, California would become the first U.S. state to enact such legislation, following similar laws in 13 other countries.