AI, Quantum Can Help Address Increasingly Complicated Security Landscape: AT&T

AI, quantum science and other emerging technologies can make telecom networks more secure, but they also create new challenges when used by bad actors, Rich Baich, AT&T chief information security officer, said during an AT&T forum Wednesday. Former Cybersecurity and Infrastructure Security Agency Director Christopher Krebs said the government will always be limited in the role in can play in making networks more secure. The forum comes as the FCC considers a notice of inquiry on using AI to curb unwanted robocalls (see 2310250070).

“It’s not so much the science, it’s the application of the science that creates the risks,” said Baich, who formerly headed cybersecurity at the CIA. “There’s this rapid pace of change associated with all these emerging technologies, which causes us to think differently about these cybersecurity issues,” he said. AI and the power of quantum computing can help rapidly sort through data and use the network to detect problems, he said.

Cybersecurity is now a board-level topic at companies like AT&T, said Rita Marty, AT&T vice president-network security. Investments in cybersecurity increase each year, but the costs of breaches also are going up, she said. Big breaches can cost hundreds of millions of dollars, she said. Nation-states now present the biggest threats as “it’s no longer rogue actors,” she said.

“Security is the No. 1 concern for all our customers, and it comes up in every conversation” with customers big and small, said Senthil Ramakrishnan, AT&T assistant vice president-technology security. The challenge is the complexity of the cybersecurity ecosystem with more than 2,000 vendors selling thousands of products, he said.

Most companies aren’t able to sort through every option that’s available, pick the right products and deploy them, Ramakrishnan said. “It’s a big challenge, but customers aren’t ready to solve it” and look to companies like AT&T for help, he said. Customers also want simplicity, he said. “The internet is becoming the new copper network,” he said. Small and mid-sized companies “are completely unprepared and underserved today,” he said: “There are a lot of security solutions, but they’re all very focused on the upper end of the market,” he said. Customers end up with a “super-complicated solution,” which they don’t have the staff to implement properly, he said.

“The cybersecurity ecosystem is very fragmented” with “many, many players,” Marty said. The average business has to deploy about 70 security tools and the tools don’t work together, she said.

The number of companies critical to U.S. IT infrastructure has grown markedly, beyond big players like AT&T, said Krebs, now a consultant. The U.S. government can’t compete with the private sector on security, he said: “The innovation cycles just aren’t there. The costs are going to be too high.”

Even having a U.S. government “threat feed” won’t work as well as something from industry, Krebs said. If the government had the appropriate information, it would be declassified and shared, he said. “There’s not this huge trove of information that’s actionable, ready to go,” he said. The private sector is able to “pull together the exact same indicators” available to the government more quickly at a much lower cost, he said.

The government has a role to play in information sharing and bringing people together, “cutting through antitrust, cutting through monopoly restrictions,” Krebs said.

Generative AI

Executives also discussed AT&T’s increasing use of generative AI (see 2306200020), an initiative launched a year ago. “AI is not new to AT&T -- AT&T was kind of there when the term artificial intelligence was founded,” said Andy Markus, AT&T chief data officer. What AT&T is doing “is kind of the next phase,” he said. Markus said AT&T doesn’t have that many employees working on generative AI, but it’s being rolled out for the entire company.

AT&T decided it would launch its own generative AI, Ask AT&T, “with equal or better functionality” than what’s available outside the company and allow employees to have access, based on their jobs at the company, Markus said. “We have our most important information as a company exposed here. We want to make sure that the right people have access to the right information,” he said.

AT&T had to create AI policy tied to its employee code of conduct, Markus said. “This was going to touch everybody, not just data scientists,” he said. Privacy has been a “stumbling block” for some companies, he said. Mark Austin, AT&T vice president-data science, said the company’s AI has already identified 500 use cases.