I don’t know where you are, but the data analysis of the RSA Conference by the prestigious Cyentia Institute is amazing. They wrote algorithms to tell us what the “most important” talks are each year from 25 years of security conference data, and illustrate our industry’s trend over time. Who can forget “A top 10 topic in 2009 was PDAs”?

This is the slide that made everyone laugh, of course:

Trends going up? GDPR, Ransomware, Financial Gain and Extortion. Big Data exploded up and then trends down over the last five years.

Trends going down? BYOD, SOX, GRC, Hacktivism, Targeted Attack, Endpoint, Mobile Device, Audit, PCI-DSS, APT, Spam…

Endpoint going down is fascinating, given how a current ex-McAfee Marketing Executive war is going full-bore. RSAC 2018 Expo Protip: people working inside Crowdstrike and Cylance are hinting on the show floor how unhappy they are with noise made about a high-bar of attribution to threat actors given their actual product low-bar performance and value.

That’s just a pro doing qualitative sampling, though. Who knows how reliable sources are, so consider as well the implication of qualitative analysis.

Some cyber companies talk threat actor in the way that Lockheed-martin talks when they want to sell you their latest bomb technology. Is that bomb effective? Depends how and what we measure. Ask me about 1968 OP IGLOO WHITE spending $1B/year on technology based on threat actor discussions almost exactly like those we see in the ex-McAfee Marketing Executive company booths…

Welcome to SF everyone! As the RSA Conference week begins, which really is a cluster of hundreds of security conferences running simultaneously for over 40,000 people converging from around the world, I sometimes get asked for local curiosities.

As a historian I feel the pull towards the past, and this year is no exception. Here are three fine examples from hundreds of interesting security landmarks in SF.

Chinese Telephone Exchange

During a period of rampant xenophobia in America, as European immigrants were committing acts of mass murder (e.g. Deep Creek, Rock Springs) against Asian immigrants, a Chinese switchboard in 1887 came to life in SF (just before the Scott Act). By 1901 it moved into a 3-tier building at 743 Washington Street. Here’s a little context for how and why the Chinese Telephone Exchange was separated from other telephone services:

Today when you visit Chinatown in SF you may notice free tea tastings are all around. This is a distant reminder of life 100 years ago, even for visitors to the Chinese Telephone Exchange, as a San Francisco Examiner report describes in 1901:

Tea and tobacco are always served to visitors, a compliment of hospitality which no Chinese business transaction is complete

At it’s peak of operation about 40 women memorized the names and switching algorithms for 1,500 lines in five dialects of Chinese, as well as English of course. Rather than use numbers, callers would ask to be connected to a person by name.

The service switched over 13,000 connections per day until it closed in 1949. Initially only men were hired, although after the 1906 earthquake only women were. Any guesses as to why? An Examiner reporter in 1901 again gives context, explaining that men used anti-competitive practices to make women too expensive to hire:

The Chinese telephone company was to put in girl operators when the exchange was refitted, and doubtless it will be done eventually. The company prefers women operators for many reasons, chiefly on account of good temper.

But when the company found that girls would be unobtainable unless they were purchased outright, and that it would be necessary to keep a platoon of armed men to guard them, to say nothing of an official chaperon to look after the proprieties, the idea of girl operators was abandoned.

“They come too high,” remarks the facetious general manager, “but in the next century we’ll be able to afford them, for girls will be cheaper then.”

Pacific Telephone Building

One of the first really tall developments in SF, which towered above the skyline (so tall it was used to fly weather warning flags and lights) for the next 40 years, were the Pacific Telephone offices. At 140 Montgomery Street, PacTel poured $4 million into their flagship office building for 2,000 women to handle the explosive growth of telephone switching services (a far cry from the 40 mentioned above at 743 Washington Street).

By 1928, the year after 140 New Montgomery was completed, the San Francisco Examiner declared “with clay from a hole in the ground in Lincoln, California, the modern city of San Francisco has come.”

It was modeled after a Gottlieb Eliel Saarinen design that lost a Chicago competition, and came to life because of the infamous local architect Timothy Pflueger. Pflueger never went to college yet left us a number of iconic buildings such as Olympic Club, Castro Theater, Alhambra Theater, and perhaps most notably for locals, a series of beautiful cocktail lounges created in the prohibition years.

AT&T Wiretap

Fast-forward to today and there are several windowless tall buildings scattered about the city, filled with automated switched connecting the city’s copper and fiber. One of particular note is 611 Folsom Street, near the latest boom in startups.

Unlike the many years of American history where telco staff would regularly moonlight by working for the police, this building gained attention for a retired member of staff who disclosed his surprise and disgust that President Bush had setup surreptitious multi-gigabit taps on telco peering links.

“What the heck is the NSA doing here?” Mark Klein, a former AT&T technician, said he asked himself.

A year or so later, he stumbled upon documents that, he said, nearly caused him to fall out of his chair. The documents, he said, show that the NSA gained access to massive amounts of e-mail and search and other Internet records of more than a dozen global and regional telecommunications providers. AT&T allowed the agency to hook into its network at a facility in San Francisco and, according to Klein, many of the other telecom companies probably knew nothing about it.

[…]

The job entailed building a “secret room” in an AT&T office 10 blocks away, he said. By coincidence, in October 2003, Klein was transferred to that office and assigned to the Internet room. He asked a technician there about the secret room on the 6th floor, and the technician told him it was connected to the Internet room a floor above. The technician, who was about to retire, handed him some wiring diagrams.

“That was my ‘aha!’ moment,” Klein said. “They’re sending the entire Internet to the secret room.”

[…]

Klein was last in Washington in 1969, to take part in an antiwar protest. Now, he said with a chuckle, he’s here in a gray suit as a lobbyist.

In some sense we’ve come a long way since 1887, tempting us to look at how different things are from technological change, and yet in other ways things haven’t moved very far at all.

In a nutshell, traditional definitions of war linked to kinetic action and physical space are being framed as overly restrictive given a desire by some to engage in offensive attacks online. The head of NSA is asking whether reducing that link and authorizing cyber attack within a new definition of “war” would affect the “comfort” of those holding responsibility.

“[On offense] the area where I think we still need to get a little more speed and agility — and as Mr. Rapuano indicated it is an area that is currently under review right now — what is the level of comfort in applying those capabilities outside designated areas of hostility,” Rogers asked out loud.

“I don’t believe anyone should grant Cyber Command or Adm. Rogers a blank ticket to do whatever you want, that is not appropriate. The part I am trying to figure out is what is the appropriate balance to ensure the broader set of stakeholders have a voice.”

Rapuano also referenced challenges associated with defining “war” in the context of cyber, which can be borderless due to the interconnected nature of the internet.

“In a domain that is so novel in many respects, and for which we do not have the empirical data and experience associated with military operations per say particularly outside areas of conflict, there are some relatively ambiguous areas around ‘well what constitutes traditional military activities,'” said Rapuano. “This is something that we are looking at within the administration and we’ve had a number of discussions with members and your staffs; so that’s an area we’re looking at to understand the trades and implications of changing the current definition.”

While I enjoy people characterizing the cyber domain as novel and border-less, let’s not kid ourselves too much. The Internet has far more borders and controls established, let alone a capability to deploy more at speed, given they are primarily software based. I can deploy over 40,000 new domains with high walls in 24 hours and there’s simply no way to leverage borders as effectively in a physical world.

Even more to the point I can distribute keys to access in such a way that it spans authorities and bureaucratically slows any attempts to break in, thus raising a far stronger multi-jurisdictional border to entry than any physical crossing.

We do ourselves no favors pretending technology is always weaker, disallowing for the prospect of a shift to stronger boundaries of less cost, and forgetting that Internet engineering is not so much truly novel as a revision of prior attempts in history (e.g. evolution of transit systems).

My recent talk at AppSecCali for example points out how barbed wire combined with repeating rifles established borders faster and more effectively than the far more “physical” barriers that came before. Now imagine someone in the 1800s calling a giant field with barbed wire border-less because it was harder for them to see in the same context as a river or mountain…