GSA scraps $15B Alliant 2 small business contract

The General Service Administration has canceled the solicitation of its $15 billion Alliant 2 Small Business governmentwide IT contract after it was plagued with protests.

That agency announced late Thursday it is “planning a new approach” for small business IT governmentwide acquisition contracts (GWACs) after “the federal government’s requirements have evolved and GSA recognizes the opportunity to strengthen, innovate, and better respond to changing technology needs and security threats,” it vaguely said in a release.

GSA had originally awarded the 10-year contract to 81 small businesses in 2018. But just over a year later, the agency rescinded those awards after one company’s successful bid protest — there were dozens of protests of the contract in total — and went back to evaluating bids.

A “new and enhanced” GWAC for small business will better reflect “the changing landscape and addresses the needs of agencies to access the expertise of small businesses” and better support “recent developments in cybersecurity, emerging technologies, and supply chain risk management,” GSA said.

“The needs of our customer agencies, small business partners, and industry partners are rapidly evolving,” Laura Stanton, GSA acting assistant commissioner for the Office of Information Technology Category, said in a statement. “GSA is committed to finding ways for our GWACs to reflect the current IT marketplace so that we can maximize the opportunities for small and women-owned, HUBzone, service-disabled veteran-owned, and 8(a) small businesses to contract with the government for cybersecurity, emerging technology, and IT supply chain risk management needs. We are working to expand the number of master contract awards to highly qualified small businesses on our GWACs, while focusing on technology requirements that support our customer agencies for future mission success.”

The first reflection of the new strategy is the solicitation of bids under the 8(a) STARS III GWAC later in July. GSA says since issuing the draft solicitation last summer, it has incorporated much of the feedback it received for STARS III, “with an even greater focus on supply chain security, emerging technologies, and performance outside of the continental United States.”

Until then, GSA suggests agencies look to other existing small business contracts, like STARS II — with a recently raised ceiling of $22 billion — VETS 2 and the IT category of the consolidated Multiple Award Schedule.

With a pandemic and hurricane season crushing FEMA, the agency could use some bots

The Federal Emergency Management Agency‘s various grant programs have made more awards more this year than the past 30 years combined, a top official says, and now the agency is considering how robotic process automation (RPA) bots could help improve the payment process.

“From a business and management stand point, automation would be a game-changer,” Chief Financial Officer Mary Comans said Wednesday during the IBM Think Gov digital event, produced by FedScoop.

FEMA awards many types of grants, including emergency preparedness funds to state, local, tribal and territorial governments. Automating detail-oriented and medial tasks away from manual labor would help the government respond to crises such as the coronavirus pandemic and major storms. The summer only gets busier for the agency after hurricane season begins June 1.

RPA bots could help free up time for FEMA workers to concentrate on more critical decisions, Comans said.

The pandemic presents unique challenges to FEMA, both internally and externally. With emergency declarations covering the country and FEMA coordinating with 40 other federal agencies, moving to maximum telework only added to the agency’s challenges, she said. With employees out of the office, automating some of their work would alleviate strained resources.

Comans said her primary responsibility was the “health and safety of the workforce.” While RPA bots won’t solve everything, she said being able to automate more tasks would be extremely helpful in the agency’s modernization journey.

“The field of emergency management needs to evolve,” she said.

Among the ideas for bots are validating upfront eligibility for grant recipients or streamlining the process for assigning funding. Automation also would catch any accidental overpayments or payments to the wrong organization, Comans said.

“We need to ensure that at the end of the day every dollar goes to the survivor that needs it,” she said.

Other agencies, like the General Services Administration, have also turned to bots as a coronavirus-response tool. One of GSA’s RPAs compiled infection data from countries where federal buildings are located. The data helped inform the government’s situational awareness and the potential risk of infection for federal workers.

Groups like the Defense Innovation Unit have also prioritized automation technology more broadly, seeking machine learning solutions to advance the complexity of the tasks bots can do.

Maria Roat: Government must become more digital to respond during crises like COVID-19

The coronavirus pandemic made apparent the need for government to expand its digital footprint to serve citizens in a more streamlined, modern fashion, said Deputy Federal CIO Maria Roat.

“I think the need to be digital was even more evident than then people anticipated, right?” Roat said Wednesday at the IBM’s Think Gov digital event, produced by FedScoop. “Everybody’s got a mobile phone. Everybody expects that digital interaction, not just with their friends and family but with their government. So I think this really put a big spotlight on that and where there’s a need for more digital enhancements and how we interface in that customer experience with every person that we deal with across the country.”

While many agencies, like the Department of Veterans Affairs, were able to rise to the occasion to improve digital interactions with Americans during the pandemic, Roat pointed to necessary opportunities to improve “authentication and identity management as we interface with our citizens” as takeaways from the crisis.

“Those are some of the lessons learned and challenges that we bring forward and how do we address those as the government continues to expand its digital footprint,” she said.

Collaboration tools

Roat pointed to agencies’ use of internal collaboration tools as a bright spot during the pandemic. At the Small Business Administration, where Roat was CIO until recently, those collaboration tools were already turned on — but their usage spiked during the move to mass telework, she said.

“I could see the usage, I had data and I could see the charts, and I could see the usage and how it grew,” Roat said.

But it could be better, she said. Particularly, collaboration tools could be used across government among agencies.

“I’d like to see more across government usage of the collaboration capabilities,” Roat said. “So the same things I was able to do within my agency — why can’t I do that with another agency? I’m at the Small Business Administration, I’m interfacing with Treasury and interfacing with all these other agencies.”

The issue isn’t the technology, but the “the processes and the policies and records management and all of those things that go with it,” she said. “So I’d like to see that continue and that work around that.”

Data sharing

Roat also celebrated success around data sharing during the pandemic, attributing it to the development of the Federal Data Strategy.

“The pandemic shows even more the need for data and how to use data and just the importance of that,” she said.

Roat pointed to the Department of Health and Human Services‘ Palantir-based HHS Protect platform as an example. “In nine days, they created a data capability where they linked 187 data sets from federal agencies, from the states, health care facilities, from academia and others. And they brought all that together. That’s just one example of the data piece and the need to continue to accelerate the use of data. And I think the Federal Data Strategy really set a lot of the foundation and a long term strategy for the federal government and its use of data.”

With any such success, it’s important to sustain that momentum after the pandemic eventually subsides, Roat said.

“There were so many things that happened so fast, and people were able to cut through and really make things happen to respond very quickly to innovate and implement changes,” she said. “I want to see that continue. And for that to continue, there needs to be that focus on long-term modernization across the federal government.”

Roat said there’s been a dialogue with the Federal CIO Council on what needs to be done to make sure that acceleration continues.

“So it’s not, you know, a one and done,” she said. “We went through the pandemic, we’re still teleworking. But for the long term, what are those things we need to continue to put resources behind as a community?”

National security-focused Anduril raises $200M in venture funding

Anduril Industries, a technology startup focused on earning national security and government work, continues to raise private funding with a Series C round of $200 million, the company announced Wednesday.

The multi-million-dollar haul from investors could illustrate a modern-day shift in Silicon Valley, a place with historical connections to serving American national security but has grown more recently skeptical of working with agencies like the Department of Defense for both ethical and financial reasons. Even major technology companies like Google have faced internal backlash for connections to military projects.

Anduril develops both hardware and software solutions that leverage artificial intelligence and has been involved in early contracts on the Air Force’s Advanced Battle Management System. Its website also lists the Customs and Border Patrol and Northern Command as other U.S.-based partners. The latest round of investments nearly doubles the company’s post-money value to $1.9 billion, according to Anduril.

“We founded Anduril because we believe there is value in Silicon Valley technology companies partnering with the Department of Defense,” co-founder and CEO Brian Schimpf said in a statement. “In just three years we have demonstrated there is a desire for real innovation in defense.”

Schimpf is a former senior employee at Palantir, another technology company that has embraced government work with some controversy. Like Palantir, Anduril also had initial backing from controversial technology executive Peter Thiel. Anduril also hired the former staff director of the Senate Armed Services Committee, Chris Brose, who now works as Anduril’s chief of strategy.

Despite the latest round of funding, challenges still remain for companies like Anduril to earn long-term business in the national security space. There is what is known as the “valley of death” for medium-sized contractors who can easily access early, small seed-like funding from the government but struggle to jump into the ring of big, prime contracts often awarded to established corporations.

Long contracting timelines and slow cybersecurity clearances also create barriers for companies like Anduril trying to sell code and tech. For now, Anduril aims to charge headfirst into the valley and hope to come out the other side alive.

“Our business continues to accelerate with a suite of software and hardware products and an expanding roster of customers,” Schimpf said.

CBP looks to standardize its hybrid cloud environment

U.S. Customs and Border Protection needs more resources to standardize its hybrid cloud environment so it can move applications anywhere on the network.

The agency’s environment is further along than most because it’s spent the last couple years working with IBM on mainframe-as-a-service, where the vendor provides both cloud and data center storage, for an annual savings between $13 million and $15 million.

Standardizing cloud computing will improve interoperability — allowing CBP to use more application programming interfaces in lieu of one-off builds — and security of continuous integration and continuous delivery pipelines, said Ed Mays, executive director of the Enterprise Data Management and Engineering Directorate at CBP.

“Ultimately our broad goal for hybrid cloud is to be able to build once and migrate anywhere based on the economic conditions at the time and based on the ability to deliver capability to our customers,” Mays said Wednesday during IBM‘s Think Gov digital event, produced by FedScoop.

CBP needs to be able to do that in an “agile, secure and standardized manner” and “can do a little bit of all of it today,” but the rest will take an “inordinate amount of resources,” he added.

For that reason, the agency has turned to industry for solutions that speed up the work while saving money.

Advanced automation and machine learning tools will be critical to performing tasks that free up CBP’s workforce in areas like passenger travel, border security, commerce, and fruit inspections, Mays said.

“There are challenges there where artificial intelligence can do things much better than a human being,” he said.

The DOD needs to define AI and protect its data better, watchdog says

What is artificial intelligence, anyway?

It’s a question that the Department of Defense should answer, according to a new report by DOD’s inspector general. The watchdog says that while parts of the DOD have their own definitions, the department must settle on a standard, establish strong governance structures for the technology and develop more consistent security controls so as not to put the military’s AI technology and other systems at risk.

“Without consistent application of security controls, malicious actors can exploit vulnerabilities on the networks and systems of DoD Components and contractors and steal information related to some of the Nation’s most valuable AI technologies,” the report states.

The desired security controls appear to be basic, like using strong passwords and monitoring for unusual network activity. Much of the security updates need to happen at service-level offices working on AI, but contractors also must be included in the uniform standards as well, the IG says.

Need for coordination

The report commends the DOD’s early work to adopt goals and initiatives, and incorporate ethics principles into its AI development. But more standardization of that work needs to happen for it to mean something, the IG says. Much of the department-wide standardization and coordination needs to happen in the Joint AI Center (JAIC), the DOD’s AI hub.

“As of March 2020, while the JAIC has taken some steps, additional actions are needed to develop and implement an AI governance framework and standards,” the report said.

Much of the IG report echos criticism from a RAND Cooperation report on the JAIC. The RAND report detailed a lack of structure in the new office and recommend better coordination across the department, as does the IG report.

Responding to the report, the DOD CIO said that the JAIC has taken several steps already that the IG recommend. They include plans for a “AI Executive Steering Group” and several other working groups and subcommittees to coordinate work in specific areas like workforce recruitment and standards across the departments.

“The final report does not completely reflect a number of actions the JAIC took over the past year to enhance DoD-wide AI governance and to accelerate scaling AI and its impact across the DoD,” the CIO wrote to the IG.

The fuel that all AI runs on — data — is still in short (usable) supply. The IG recommended the DOD CIO set up more data-sharing mechanisms. While data sharing will increase the ability for data-driven projects to flourish, the JAIC needs better visibility as to how many AI initiatives are underway across the department.

Currently, the DOD doesn’t know how many AI projects its many components have under way. That’s a problem if offices like the JAIC are to be a central hub for both AI policy and fielding.

“Without a reliable baseline of active AI projects within the DoD … the JAIC will not be able to effectively execute its mission to maintain an accounting of DoD AI initiatives,” the report stated.

Pandemic pushes JAIC to continue AI humanitarian efforts

Even as the Joint Artificial Intelligence Center expands its warfighting work, the still-growing office is still focused on several humanitarian relief projects as the coronavirus pandemic continues to disrupt life in the U.S.

The first sprint the JAIC did in response to the pandemic was to give National Guard and Northern Command service members a dashboard that could monitor supply chain shortages and panic-buying. The tools, made under the moniker Project Salus, went from “idea to code” in two weeks, according to project lead and missions director at the JAIC, Col. Robert Kinney. The project continues to grow with the JAIC planning to deliver six more models to the guard on top of the 10 already in use.

“This is has been a fantastic effort,” Kinney said Wednesday about Project Salus during the IBM Think Gov digital event, produced by FedScoop.

While the Department of Defense established JAIC ultimately to handle technology related to the battlefield, most of its early projects were lower-risk humanitarian missions like tracking wildfires.

Other humanitarian work by the JAIC includes a search-and-rescue system. The technology, in its early stages, would help search for lost members of the military and civilians both on land and at sea. Currently, search and rescue is a time-intensive manual process that uses binoculars as a principle piece of tech. Some of the process could be automated with computer vision and advanced analytics.

Project Salus might see its code be repurposed in other ways as well. JAIC spokesperson, Lt. Cmdr. Arlo Abrahamson, told FedScoop on Wednesday that separate from the national COVID-19 response, the JAIC is in talks with Indo-Pacific Command on humanitarian work inspired by Salus’s analytic capabilities.

JAIC wants “to determine if aspects of the Project Salus capability and predictive analytics could apply to potential humanitarian assistance and disaster relief missions supported by the DoD,” he said in an email.

Kinney said that the JAIC’s humanitarian assistance work is “near and dear to my heart” and a huge motivator. The JAIC recently amped up work and spending on its warfighting mission initiative and its now-departed first director said that its time to bring AI to the battlefield.

“The JAIC is a game changer and every mission area that we concentrate on has just great potential,” Kinney said.

Microsegmentation isn’t a zero trust ‘easy button,’ but it’s helping agencies telework

Agencies eager to build zero-trust security architectures that support increased telework are finding microsegmentation isn’t a quick fix.

Coronavirus-related telework has seen agencies more quickly adopting use cases under the Trusted Internet Connections 3.0 — updated guidance for securing networks that adjusts to advancements in cybersecurity since TIC 2.0 came out in 2007.

TIC 3.0 encourages segmenting networks into “trust zones,” elastic groups of assets with similar protection requirements, but that’s only part of zero trust, said Sean Connelly, TIC program manager at the Cybersecurity and Infrastructure Security Agency.

“We are starting to talk to a number of agencies to look at microsegmentation architectures. It’s one step toward that modernization of their environment,” Connelly said Wednesday during IBM‘s Think Gov digital event, produced by FedScoop. “We look at it as a way for them to secure their traffic across the enterprise, as opposed to focusing on the traditional network perimeter.”

But one unnamed federal research agency recently sought microsegmentation as the foundation for its zero-trust architecture, when it wouldn’t solve the problem at hand, said Aarti Borkar, vice president of IBM Security. The agency regularly collaborates with international groups that need access to systems containing research data — access that would be jeopardized.

“Everybody’s looking for the easy button for zero trust,” Borkar said. “And vendors help cause that confusion because we’re looking at it from one vantage point, one piece.”

Most agencies should prepare for zero trust to coexist with their traditional network security strategies for some time because they’re still dealing with on-premise data centers, hybrid networks or multi-cloud networks, Connelly said.

Microsegmentation is compatible with zero trust, but agencies need a detailed inventory of their assets, users and, in particular, business processes to truly understand their workflows well enough to build out security policies.

“Right now zero trust is not a complete architecture, where a lot of agencies are in a position to adopt zero trust across the entire enterprise,” Connelly said.

As the pandemic rages on in the U.S., agencies have adopted entirely new information technology platforms on outside networks to enable mobile and bring-your-own-device work.

The TIC program is distilling lessons learned from such rollouts into a zero-trust use case that will provide agencies with security patterns and capabilities to consider, Connelly said.

“So it’s going to take a while I think before you start to see a majority of agencies support zero trust,” Connelly said. “But I think a lot of those technologies are there.”

Telework shift presented durable lessons about workforce, USPS and NASA officials say

After months of adjusting to an era of mass telework, agency officials say they are still wrapping their heads around some of the lessons from that process — and not just the stumbling blocks, but also the successes.

Greg Crabb, the chief information security officer for the U.S. Postal Service, says the pandemic has given him fresh perspective on what he calls the “four Cs” of innovation: compete, collaborate, control and create. The competition, this time, was about staying ahead of cybersecurity threats, not necessarily competing with other businesses, he says. Collaboration lessons were direct, too: The USPS had to rethink its relationships with other supply chain companies like Amazon and FedEx.

The lessons in the “control” and “create” areas were tied directly to the workforce, Crabb says.

“We moved literally tens of thousands of employees off the workroom floor, into their homes in order to be able to support the frontline workers,” Crabb said Wednesday during IBM‘s ThinkGov digital event, produced by FedScoop. “And for those frontline workers we had to create new ways of interacting,” and the net result is that Americans didn’t really see any drop-off in service.

NASA Chief Human Capital Officer Jane Datta says the coronavirus shutdowns caused the space agency to reconsider what kinds of people it needs for jobs, and where they do those jobs.

“We are really going to leverage what we’ve learned over these last months, in remote work,” Datta said during the IBM event. “Because the more you can have flexibility on geography, and people working from where they are, the greater access to talent we have … as well as plugging people we already have working for NASA into work, wherever that work might be.”

The Postal Service made a few early decisions that spotlighted tech talent as the organization shifted to mass telework, Crabb said. One was creating a “COVID command” for the entire agency, with the CIO Kristin Seaver in charge. Within his own group of about 600 cybersecurity workers, Crabb said he formed a “tiger team” to focus on information security.

“I put some of my up-and-comers on that team, so that they could shine,” he said. “That’s extremely important: finding somebody that can really handle a situation under pressure.”

Going forward, organizations will need to define the “new normal,” Crabb said. Every company will have to decide if it still needs employees working side-by-side.

“We’ve learned a lot of lessons” at the Postal Service, he said. “And the first lesson is that we need to keep our employees safe.”

By using AI, the VA dramatically decreased claims processing intake times, official says

The Veteran Benefits Administration was spending too much time processing its mail. So it turned to artificial intelligence-enabled systems to slash its processing time.

By applying AI that could more quickly sort incoming claims from multiple inputs — mail, fax and electronic submissions — the Department of Veterans Affairs was able to reduce the time just sorting the claims from 10 days to about half a day, Paul Lawrence, the head of the VBA, said during the IBM Think Gov digital event, produced by FedScoop.

“We realized we had far too many people doing old-fashioned manual labor,” Lawrence said.

Veterans send in documents to the VA that pertain to the type of benefits they are eligible for and to settle disputes over services. More than 5 million veterans file claims, and the current backlog runs more than 150,000, a number that has dropped drastically in recent years.

Now that sorting times are reduced with the assistance of technology, Lawrence said the VBA is focused on other workflow challenges. He wants to reskill employees away from manual, tedious tasks and have them work with technology that can help sort and leave the complex decision making to humans. Much of the work around settling claim disputes and making decisions about benefits is not something AI has the ability to tackle.

The initial successes of technology have not solved all of the VBA’s challenges, however. The administration still takes almost 100 days to process claims on average, according to its website.

Lawrence also wants to be able to be more transparent about where claims are in the process, much like how pizza delivery services can now show customers where their orders are in the cooking and transport process, Lawrence said.

The lessons learned from using AI to improve initial processing times will stick around, he said.

“It is now something we are really proud of, we are happy to talk about how good we are,” Lawrence said.