It goes fast! Even if your bank account has more than $260 billion sitting in it, Matt writes.

The big picture: Wall Street analysts are keeping an increasingly close eye on the U.S. Treasury's published finances, as they try to estimate when the Federal government could run out of money if the debt ceiling isn't lifted.

Context: The U.S. reached its statutory debt limit of $31.4 trillion in January, constraining the Treasury's ability to borrow.

• The government can get along all right for a while via "extraordinary measures" to ensure it meets its obligations. That means deferring payments to some government pension plans and using up its existing cash balances, among other financial tweaks.
• But as time goes on it gets harder and harder.

State of play: Analysts have been watching recent tax receipts especially closely, hoping that a bumper crop of tax payments could push out the day of reckoning — also known as the "X-date," when the government runs out of money and could default on payments. But no dice.

• It's true that that balance at the Federal government's checking account at the Federal Reserve — the Treasury General Account, charted above — got a bump over the last few days.
• But analysts have noted that tax receipts have been coming in lower than anticipated, perhaps partially due to a drop in capital gains taxes given the awful year the markets had in 2022.

What they're saying: "April tax receipts have been on the low side of our expectations and the deficit is tracking higher," analysts with BofA Securities wrote in a note on Friday. They pulled forward their X-date estimate to early August (from mid-August), and added that it "could easily slip to late July."

• Separately, Goldman Sachs analysts wrote that their "base case remains for the debt limit deadline to fall in late July. However, the odds of an early June deadline have edged up."
• For the record, the Treasury Department itself has said it will start running out of wiggle room in early June.

The bottom line: The pressure on the White House and congressional Republicans to reach some sort of deal could get serious pretty quickly.

🔋 GM, Samsung plan new U.S. battery plant. (AP)

🏦 UBS reported strong customer inflows ahead of Credit Suisse integration. (WSJ)

📺 Sudden firings signal a new era for cable news. (Axios)

The riptide of client money out of First Republic was even stronger than analysts expected, Matt writes.

Driving the news: The San Francisco-based lender, which was battered by the collapse in confidence that followed Silicon Valley Bank's failure, reported that its deposits plunged by nearly $72 billion during the quarter that ended in March.

• The company disclosed the drop — which amounted to more than 40% of its deposits — in its quarterly earnings report shortly after the close of trading Monday.
• That 40% decline comes despite attempts at an industry-led rescue, in which major banks stashed $30 billion in deposits at First Republic in an effort to shore up confidence.

What they're saying: "The situation is worse than even bears had thought, in our view," wrote David Smith, an analyst with Autonomous Research, shortly after the release of the report.

• In its statement, the bank indicated that the outflow of deposits had relented since late March.

Between the lines: The market didn't like the news, and the stock dropped about 20% in after-hours trading.

• It's down over 90% in the last 12 months — making it the single worst-performing company in the S&P 500 over that period.

Productivity improved considerably within the customer service department of a Fortune 500 company that used generative AI, per a new study from researchers at Stanford and MIT, Emily writes.

Why it matters: The researchers believe it's the first empirical evidence of the effects of generative AI — artificial intelligence that creates (generates) content, like text or images — on the workplace.

What they did: Researchers tested AI software with a customer service team of more than 5,000 agents at an unnamed Fortune 500 company that provides software to small businesses.

• The AI monitored customer chats and gave agents real-time suggestions for how to respond, including ideas for wording — key to keeping customers from growing hostile — and links to technical information to help troubleshoot issues.
• Agents were free to ignore the advice.

What they found: Use of the AI led to a 14% increase in the number of customer service chats an agent successfully responded to per hour.

• Agents spent less time handling individual chats, and were able to take care of more customers per hour — there was also a small increase in the share of chats resolved successfully.
• The AI had the biggest impact — and helped reduce turnover — among the lowest-skilled customer service agents new to the job. Experienced customer service agents saw only a slight lift.
• That's because they already have learned the information the AI was sharing — in this case, the AI basically serves as a way for more experienced customer service agents to transfer knowledge to newbies.

The bottom line: This is just one study of a single workplace, where the product (the software) doesn't really change much. Many more studies of more workplaces are needed to truly understand its impact.

Americans with a bit of cash to spare could earn more than 4% interest on a plain vanilla high-yield savings account right now — but few are taking advantage of the opportunity, Emily writes.

Driving the news: Only 7% of Americans with short-term savings are earning that much, per a Bankrate survey last month.

• Only around 1 in 5 said they’re earning a yield that’s 3% or higher.
• The average rate consumers are earning on their savings is a paltry 0.24%, Greg McBride, chief analyst at Bankrate, told Axios.

The big picture: This is good news for banks that make more money when customer cash languishes at low rates, as Axios' Felix Salmon wrote last weekend.

• It's not great for savers, especially at a time when inflation is at 5% — and money is losing value if it's just sitting there earning very little.

Separately, 33% of those surveyed had no savings at all — a whole other problem.