The Kitchen Sync
§230 – Mend it, don’t end it
What’s new: Heritage says Section 230 needs an update.
Why this matters: Section 230 of the Communications Decency Act (CDA) has been critical to the development of today’s Internet and Internet services. But the expanding presence of these services and a growing political distrust of “Big Tech” require the scope and language of Section 230 to be refined.
Key points: On Tuesday, we published “Section 230 — Mend It, Don’t End It.” Here’s what you need to know.
Congress should refine Section 230 of the CDA to ensure our markets and our civil discourse remain free and fair.
We believe Google, Facebook, Twitter, and others have squandered the public trust and are now rightly being scrutinized for their inconsistent — and often political — moderation of user content.
The Heritage Foundation has been deliberate in the development of these recommendations and they are carefully crafted to refine the scope and language of Section 230 to better fit the statute’s original intent and to restrain potential abuses of its protections.
We believe Congress, and not executive agencies, should undertake these reforms.
Even so, conservatives should be prudent in refining Section 230 so as not to provoke unintended consequences that could constrain individual and corporate speech and other freedoms.
What we’re thinking: This paper has been a long time coming and nothing about this issue is easy. This policy topic provokes real tensions between issuess like freedom of speech, private property rights, online bias, and criminal online activities. But it’s high time Section 230 was clarified and we believe this paper offers a compelling way forward. We hope you’ll give it a read.
Deep Dive: DOD investments in autonomy & AI
By Frederico Bartels, Sr. Policy Analyst
What’s new: Georgetown’s Center for Security and Emerging Technology released a report on “U.S. Military Investments in Autonomy and AI: A Budgetary Assessment”
Why this matters: It is an independent estimate of how and how much the Pentagon invests in artificial intelligence and autonomy based on data sent to Congress in budget justification books.
Context: CSET has been producing good assessments of AI spending and trends, be it in the private sector, the Chinese government, or the job market. It is common knowledge that the Department of Defense is a substantial player in the development and application of artificial intelligence – the question is how much is being invested and in what areas.
There is no budget line for “artificial intelligence” projects in the budget, so producing this estimate required going into the details of the justification books and classifying 143 PEs (program elements), 674 projects, and 2,312 components in the research, development, testing and evaluation accounts.
Go Deeper: The report highlights that DOD invests more in applied research rather than basic AI research. The applied areas are described in five different buckets:
Unmanned systems – drones, unmanned submarines
Information processing – sensor fusion, battlefield visualization
Decision support – proposal of action plans, situational understanding
Targeting functions – identification, tracking, target prioritization
Human-machine teaming – augmenting human performance, AI as a teammate rather than tool
The report sees gaps in investment in two areas: trust in human-machine teams, as in how humans understand and trust or mistrust AI; and security and safety of AI systems, as in system robustness and resilience.
Limitations: Despite the budgetary focus, the main story of the report is one of the choices being made by the Pentagon, not a financial accounting. The emphasis is on typifying and classifying the projects and activities taking place.
The total level of investment is correctly de-emphasized, because the nature of the data necessarily leads to over and underestimates by design. As the report (p.16) puts, “no easy method exists for accurately estimating U.S. military research and development investments in autonomy and AI.” However, it is the main question that is going to be asked by Congressional actors.
Future: On page 43, the report finally shows the projections for future AI spending levels until 2024 – justification books include a five-year projection for each program element. There the story is of a largely flat investment portfolio – with basic research holding steady through the period, while applied and advanced research see some drops followed by small increases.
This is materially important since projected baseline serves as the initial building block for future budgets requests.
What we’re thinking: The defense budget is organized in two basic ways: by the service or by the type of appropriations. This means that it is easy to know how much the Air Force spends on research or on personnel, but hard to quantify how much it spends on a certain type of research or how much is dedicated to the maintenance of forces in Europe. The budgetary systems need to be improved to provide different cuts of the same data. It is a weakness also highlighted by the creation of the Indo-Pacific Deterrence Initiative.
The report shows the Pentagon betting on leveraging AI to enhance autonomy of systems and process to enhance human effectiveness and sees AI as a tool rather than an independent actor. However, the dollar amount invested is still an open question to which the report offers different approximations and calculation methods, rather than one estimate.
Iran goes spear phishing in Munich
What’s new: Microsoft says Iran’s hackers are targeting the Munich Security Conference (MSC) according to CyberScoop.
Why this matters: The MSC is one of the largest and most influential security conferences held every year and regularly includes multiple heads-of-state, senior government leaders, and other geopolitical influencers.
Key points:
Microsoft’s cybersecurity researchers are warning that Iran’s state-backed “Phosphorous” hacking team is sending spear phishing emails to high-profile attendees to the upcoming MSC and the Think 20 Summit in Saudi Arabia.
The emails are made to look like they’re from conference organizers but contain malicious code and links.
More than 100 attendees have received the emails, with several high-profile attendees falling for the scam.
“We believe Phosphorus is engaging in these attacks for intelligence collection purposes,” says Tom Burt, corporate VP of Microsoft Security and Trust. “The attacks were successful in compromising several victims, including former ambassadors and other senior policy experts who help shape global agendas and foreign policies in their respective countries.”
What we’re thinking: Phish in a barrel.
Ant goes big
What’s new: China’s Ant Group will soon launch a $35 billion IPO after getting the greenlight from the nation’s government.
Why this matters: Ant Group is one of the top fintech companies and is backed by Chinese e-commerce giant Alibaba. If the projected IPO is hit, it’ll be the largest in history.
Key points:
The IPO will also be the first simultaneous listing in Hong Kong and on the year-old STAR Market in Shanghai.
Ant says it intends to split shares evenly between Hong Kong and Shanghai, selling up to 1.67 billion on each exchange.
That would represent up to 11% of its enlarged share capital, before a 15% greenshoe or over-allotment option is exercised, the prospectus showed.
But analysts are now saying buyers are struggling to secure allocations as the fundraising continues to expand.
What we’re thinking: This Ant is huge: in the last year, the company says it processed $17 trillion in digital payments in mainland China, taking in $18 billion in revenue and profits of $2.7 billion.
OSIRIS-REx successfully samples asteroid
What’s new: Last week, NASA’s OSIRIS-REx spacecraft landed on an asteroid and took samples before blasting off the rock again.
Why this matters:According to NASA, “The OSIRIS-REx spacecraft is traveling to Bennu, a carbonaceous asteroid whose regolith may record the earliest history of our solar system … Bennu is also one of the most potentially hazardous asteroids, as it has a relatively high probability of impacting the Earth late in the 22nd century.”
Key points:
OSIRIS-REx touched down on Bennu on Oct. 20 and puffed nitrogen gas at the space rock to blow pieces into the arm’s sampling head before backing away to safety.
When OSIRIS-REx scientists were able to see images of the sampling head on Oct. 22, they realized that the maneuver had been so successful that asteroid rubble blocked the flap designed to close off the material in the arm, allowing some of the precious haul to leak away.
Now, however, NASA engineers have done some sort of math magic and the problem has been fixed.
The mission is scheduled to remain at the space rock until the middle of next year before turning back to Earth, where it will deliver its cargo in 2023.
What we’re thinking: You can see some pretty sweet images from the spacecraft here.
