Major model providers are fighting for the affections of the US government.

In recent weeks, both OpenAI and Anthropic have offered access to their models to government agencies practically for free, with OpenAI proposing to provide ChatGPT Enterprise to the executive branch for $1 a year and Anthropic matching that price to supply its Claude models to all three branches of government.

As these organizations seek to weave their way into government agencies’ tech stacks, however, whether their models can safely handle sensitive private data at scale remains uncertain.

“By now, model providers are well aware of security being a big impediment to adoption,” said Arti Raman, founder and CEO of AI data security company Portal26.

Though AI models still have their kinks, the bigger security problem isn’t the tech itself but the people who are using it, she said. “The bigger risks are on the human side.”

A variety of vulnerabilities have already been demonstrated:

• According to IBM’s recent Cost of a Data Breach report, 97% of surveyed organizations that experienced AI-related security incidents reported not having access controls in place.
• Of those incidents, 60% led to compromised data, and 31% led to operational disruption.
• “Who gets access and access control becomes a bigger problem than data leaking from a model that may not be connected to the outside world … Data security risks are from person to person,” said Raman.

And in government agencies, especially those handling large amounts of sensitive information relating to civilians, the risks can be great. The organizations are often the target of threat actors already, and with the Cybersecurity and Infrastructure Security Agency facing persistent cuts under the Trump administration, the layers of protection may become even thinner.

While government workers are often “trained and conditioned to worry about security,” said Raman, the nascent and evolving nature of AI means that training and governance can’t be a static thing.

“Training and education are incredibly important,” said Raman “It can’t be in the form of a manual or something that you do once a year. It has to be done in real time.”

Education may be only part of the solution. The “white space” of AI access and identity control could represent a major opportunity in the market, said Raman. “We need some innovation on complex identity and entitlement … somebody needs to really understand how to connect the dots between what a model was trained on and has access to versus what it is and isn’t allowed to answer.”

Hey, robotaxis, we’re walking here! We’re walking here!!

Last week, Tesla took steps to bring its autonomous vehicles (AV) to the streets in New York City, following in the footsteps of Waymo, which began the process of bringing its services to the Big Apple earlier this summer. It’s just the latest move in the stop-and-go, uneven rollout of robotaxis nationwide.

On the Road

If you live in one of five US cities — San Francisco, Los Angeles, Phoenix, Austin and Atlanta — Waymo’s robotaxis may already feel like a regular part of the traffic bloodstream. The Alphabet-owned company recently claimed it’s providing 250,000 paid rides per week across the markets, way up from just a year ago. Increasingly, the company has competition: Waymo vehicles have been racing against Tesla’s robotaxi service for business in Austin since June, while Amazon’s Zoox robotaxi business hit the Las Vegas Strip last month after the company scored a crucial exemption from the National Highway Traffic Safety Administration.

Next up: conquering New York gridlock, where some 200,000 for-hire vehicles licensed by the city complete approximately 1 million trips per day, according to city data. Waymo applied for a city license in June and began a pilot testing program last month. Tesla, meanwhile, is taking a “move fast and break rules” approach familiar to its founder. Last week, the EV titan posted job listings for AV test drivers, even though the state’s Department of Transportation told Gizmodo that the company had not received permits to test AVs in the city.

The push into New York comes as other auto industry players expand their AV ambitions:

• Last week, Bloomberg reported that GM is seeking to lure back employees who had worked on its now-defunct self-driving Cruise unit, as the auto giant revives its AV project. Sources told Bloomberg the company is focusing on self-driving cars for consumers rather than robotaxis.
• In July, Uber announced a $300 million investment in EV firm Lucid, as well as an undisclosed investment in AV tech firm Nuro, with plans to bring 20,000 Lucid robotaxis equipped with Nuro tech to the roads in the next six years. Waymo offers its services through Uber’s app in Austin and Atlanta; in May, Uber CEO Dara Khosrowshahi said that Waymo AVs complete more rides per day than 99% of Uber drivers in Austin.

Fare is Fair: Still, the massive capital needed for research, development and production means “the autonomous space in the US will remain largely unprofitable until scale increases and vehicle costs materially decline,” JPMorgan analyst Doug Anmuth recently told the Financial Times. The story has been different in China, according to a Rest of World feature published last week, which said that AV players are enjoying the fruits of government spending and a top-down regulatory approach, compared to municipal control in the US. By 2030, the China Society of Automotive Engineers estimates that as many as 20% of cars sold in the country will be fully driverless.

Just like the weather, the markets change, and just like meteorologists, investors try to predict those changes.

Some industries, however, go through regular boom-and-bust cycles. Cyclical industries, encapsulating everything from airlines to raw materials, tend to track the wider economy, with expansions and contractions happening at regular intervals based on the broader market. ETFs tracking cyclicals can offer bullish investors high returns in times of prosperity, but experts said even periodic sectors can be unpredictable. “At a certain point, you’re really making bets on a single stock, or a few stocks, that are really going to be the driver of the returns,” said Dan Sotiroff, research analyst at Morningstar. “That’s what I think a lot of people miss with this stuff.”

What Goes Up…

Cyclical industries like manufacturing, consumer discretionary products, and construction typically thrive in economic expansions but suffer during slowdowns. (Americans tend to cut back on travel and discretionary spending in times of hardship.) Issuers have tried to capitalize on these patterns, carving out niche funds that promise high returns in bull markets.

Take the AdvisorShares HVAC and Industrials ETF launched earlier this year, which bundles companies that manufacture heating and cooling systems, or the Materials Select Sector SPDR Fund (XLB), which follows the materials sector of the S&P 500. (It holds just 29 stocks, with the top 10 making up over 60% of its assets.) Although short-term outlooks can look rosy, investors need to be aware of the long-term risks, Sotiroff said, particularly since cyclical and sector ETFs — unlike thematic ETFs like artificial intelligence or robotics — don’t usually span multiple industries. “[Cyclical ETFs] go through these waves where the biggest companies, the ones driving the return of the ETF, are hitting home runs, and they look great,” he said. “Then six months or a year out, they look terrible.”

Other cyclical industry ETFs include:

• The U.S. Global Jets ETF (JETS), which tracks American and international airlines, plane manufacturers, airports and airline-related communications companies.
• The Consumer Discretionary Select Sector SPDR Fund (XLY), which tracks the consumer discretionary sector of the S&P 500, including categories like entertainment, cars and retail.
• The iShares U.S. Real Estate ETF (IYR), which tracks US real estate companies and real estate investment trusts.

Fee-ble Returns. The other reason investors should be wary of these funds is that a lot of them charge “obnoxiously high” fees, Sotiroff said. (JETS has an expense ratio of 0.60% and HVAC’s is 0.89%.) “Fees dictate performance. They’re charging higher fees because they can, because they’re giving you specialized access to some niche segment of the economy,” he said. “But that tends not to work out very well.”