Advancing Public Safety and Protecting Privacy

Advancing Public Safety and Protecting Privacy

Lumina Testifies Before the Florida House of Representatives

“How do we leverage the power of technology without sacrificing constitutional liberties?  How do we ensure we are doing everything we can to keep our communities safe without turning our society into the Minority Report?  

These were the opening questions posed by Florida State Representative James Grant at a recent hearing focused on Using Technology to Advance Public Safety and Privacy in the Florida House of Representatives.

Lumina joined a panel of expert witnesses to answer these and other questions from members of the Criminal Justice Subcommittee.  

In addition to Lumina’s Doug Licker and Jessica Dareneau, other panelists included Dr. Russell Baker, CEO & Founder of Psynetix and Wayne A. Logan, a professor of law at Florida State University.

No Standardized Methodology

Beginning on the issue of using technology to keep communities safe, Psynetix’s Baker noted that the signs of violence or potential terrorism are often missed because there is no standardized methodology to collect, report and disseminate crucial information indicative of these potential acts – and that even if the data is available, it becomes siloed.

Lumina expanded on those complications, noting that 93 percent of those carrying out a mass violent attack make threatening communications prior to the event – including on social media –  and that 75 percent of terrorists used the internet to plan an attack.

The Internet is Useful to Everyone…Including Bad Actors

“The internet, it turns out is useful to everyone…and that includes bad actors,” Licker testified. 

“The UN, the FBI and the Office of the Director of National Intelligence (ODNI) support the use of new technologies to help mine the publicly available information on the internet to help prevent, predict and deter attacks in the future,” he continued.

The problem, comes in the mass amounts of data available on the web – some 2.5 quintillion bytes of data are added to the internet daily.  And, constrained resources from law enforcement agencies to analyze the data and respond.

Real-time Detection of Digital Evidence

In the slide presentation, Lumina shared a quote from the RAND Corporation which noted: “Most law-enforcement agencies in the United States, particularly at the state and local level, don’t have a whole lot of capability and technical people to manage and respond to digital evidence more generally, much less real-time detection.”

That’s where technologies like Lumina’s Radiance platform can be valuable for law enforcement.

“The power of our Radiance platform is two-fold – its ability to ingest massive amounts of unstructured, open source data and its real-time ability to analyze that information to predict and prevent organizational risks and threats,” Dareneau said. “It does this through purpose-built, best-in-class algorithms that can overcome the challenges of massive unstructured data ingestion and prioritization.”

The question of each of our publicly available digital footprints, and law enforcement’s ability to use that information in an investigation was widely discussed at the hearing.

Is Privacy Dead?

“Digital dossiers exist today on us all, which law enforcement can and will readily put to use in its work such as by means of computers, patrol cars and even hand-held devices,” Logan testified. “And why should law enforcement not be able to harness the crime control tools enabled by technological advances, such as machine learning targeting massive data sources?”

“In my view, my personal view, they should be able to do so but in a regulated manner,” he continued.

But, what should those regulations look like, and how best to approach the balance between privacy and safety?

Logan noted that the European Union, California and Illinois are all taking steps towards data protection measures, and could be models for Florida to follow.  

Transparency is Key

Dareneau said many of the policies being implemented relate to transparency.

“Transparency is so important, and that is what so many of these other jurisdictions are enacting in their legislation – requirements that you disclose what you are collecting and then how you are using it,” she testified. “So we try to stay on top of that, and make sure our privacy policy and terms includes exactly what we are collecting, how we are using it and who we could provide it to.”

As the hearing ended, Chairman Grant reiterated the work before his subcommittee to understand and delineate between private data and public information.  “This body is committed to acting,” he said.

Committed to Acting

When legislative session begins January 14, 2020, it’s clear that this topic will be a key focus for this subcommittee and the broader legislature. 

As Logan noted, “Technology is really potentially a game changer here. The question is whether it will be permitted, what limitations are going to be put on it and what accountability measures will be put in place. It’s just a different era.  We need to air the potential concerns here, and we need to transparently deliberate them and decide the issues.”

You can watch the hearing and review the materials here.

Predicting and Preventing Heath Care Fraud

Predicting and Preventing Heath Care Fraud

When the Centers for Medicare & Medicaid Services (CMS) announced its vision to modernize Medicare program integrity, Administrator Seema Verma highlighted the agency’s interest in seeking new innovative strategies involving machine learning and artificial intelligence.

Executive Order Directs HHS to use AI to Detect Fraud and Abuse

The announcement came earlier this month and followed an Executive Order by President Trump which urged the Secretary of Health and Human Services (HHS) to direct “public and private resources toward detecting and preventing fraud, waste, and abuse, including through the use of the latest technologies such as artificial intelligence.”

Medicare Fraud Estimated between $21 and $71 Billion Annually

Medicare fraud, waste, and abuse costs CMS and taxpayers billions of dollars.

In 2018, improper payments represented five percent of the total $616.8 billion of Medicare’s net costs. And it is estimated that Medicare loses between $21 and $71 billion per year to fraud, waste and abuse.

Part of those costs are driven by inefficiencies in trying to identify and flag these issues before, during and after they occur.

For example, today, clinicians manually review medical records associated with Medicare claims and as a result, CMS reviews less than one percent of those records

Artificial intelligence and machine learning could be more cost effective and less burdensome, and can help existing predictive systems designed to flag fraud.

HHS Among Largest Data Producers in the World

In order to understand the potential for AI, CMS also recently issued a Request for Information asking, among other things, if AI tools are being used in the private sector to detect fraud and how AI can enhance program integrity efforts.

HHS, which houses CMS,  is among the largest data producers in the world, with its healthcare and financial data exceeding petabytes per year, making it the perfect fit for AI and machine learning models.

In fact, researchers at Florida Atlantic University programmed computers to predict, classify and flag potentially fraudulent Medicare Part B claims from 2012-2015, using algorithms to detect patterns of fraud in publicly available CMS data.  The researchers noted they had only “scratched the surface” and planned further trials.

Just “Scratching the Surface”

But the promise of AI isn’t in just in the CMS data. It’s also in the behaviors of those looking to commit fraud.

According to Jeremy Clopton, director at accounting consultancy Upstream Academy and an Association of Certified Fraud Examiners faculty member, the risk of fraud is often described as having three key factors: a perceived pressure or financial need, a perceived opportunity, and a rationalization of the behavior.

To prevent fraud,  AI must analyze behavioral data that might indicate the pressure someone is facing and how they could rationalize fraud to deal with those pressures. For example, he notes that someone facing financial pressures might regularly search for articles related to debt relief and could also mention those concerns in emails. AI has made finding these behaviors more efficient.

AI, Fraud Detection and the Private Sector

The private sector is already embracing AI for a variety of fraud prevention needs.  Aetna has 350 machine learning models focused on preventing criminals from fabricating health insurance claims.

And, Mastercard Healthcare Solutions recently announced it would also use AI to identify suspicious activity and help its clients detect fraud .

Beyond just healthcare, the use of AI and ML as part of an organization’s anti-fraud programs is expected to almost triple in the next two years, according to the Association of Certified Fraud Examiners.  

And, 55 percent of organizations expect to increase their budgets for anti-fraud technology over the next two years.

Based on the efforts at HHS and CMS, it looks like the Federal Government will be part of the AI-fueled anti-fraud movement.

Learn more about AI-powered Radiance and its risk and fraud sensing capabilities.

Try Radiance for free today.

Ending Human Trafficking Through Education, Awareness and AI

Ending Human Trafficking Through Education, Awareness and AI

Earlier this month, Florida became the first state to require schools to teach K-12 students about child trafficking prevention. The state ranks third in the nation for reported human trafficking cases, with 767 cases reported in 2018, nearly 20 percent of which involved minors.

A $150 billion industry

While Florida’s program will be the first targeted on youth education, awareness campaigns have become a critical component of the fight against this $150 billion industry, which impacts as many as 40.3 million people annually.

The Department of Homeland Security’s Blue Campaign is one example.  This national public awareness campaign is focused on increasing detection of human trafficking and identifying victims. 

Increasing detection of victims

The campaign works to educate the public, law enforcement and industry partners to recognize the indicators of human trafficking, and how to appropriately respond to possible cases.

According to DHS, among the potential indicators that a person might be a victim of human trafficking are:

  • Disconnection from family and friends
  • Dramatic and sudden changes in behavior
  • Disorientation and signs of abuse
  • Timid, fearful or submissive behavior
  • Signs of being denied food, water, sleep or medical care
  • Deference to someone in authority or the appearance of being coached on what to say

Finding the perpetrators

While potential indicators for victims are well documented, identifying the perpetrators is more difficult.

Law enforcement points to the fact that traffickers represent every social, ethnic, and racial group and are not only men—women run many established rings.

Cases have even revealed that traffickers are not necessarily always strangers to or casual acquaintances of the victims. Traffickers can be family members, intimate partners, and long-time friends of the victims.

With all these variables in finding the perpetrators, law enforcement is increasingly looking for tools to help this lucrative and subversive crime. 

“A rare window into criminal behavior”

One tool is the Internet, which provides traffickers with the unprecedented ability to exploit a greater number of victims and advertise services across geographic areas.  It is also a way to recruit victims, especially unsuspecting and vulnerable youth. 

As research conducted in 2011 at the University of Southern California found, online trafficking transactions “leave behind traces of user activity, providing a rare window into criminal behavior, techniques, and patterns.

“Every online communication between traffickers, ‘johns,’ and their victims reveals potentially actionable information for anti-trafficking investigators.”

The study noted the potential for integrating human experts and computer-assisted technologies like AI to detect trafficking online.

AI and human trafficking

Similar research conducted at Carnegie Mellon University looked at how low-level traffickers and organized transnational criminal networks used web sites like Craigslist and Backpage to advertise their victims. The researchers developed AI-based tools to find patterns in the hundreds of millions of online ads and help recover victims and find the bad actors.

Fast forward to today.

In February, the United Nations held a two-day conference focused on using AI to end modern slavery.

The conference brought together researchers, policy makers, social scientists, members of the tech community, and survivors.

One of those researchers – from Lehigh University – is working on a human trafficking project to help law enforcement overcome the challenges of turning vast amounts of data, primarily from police incident reports, into actionable intelligence to assist with their investigations.

Providing better alerts and real risks

Former Federal government officials share the optimism about the power of AI to aid law enforcement in weeding out the criminals and finding the victims.

Alma Angotti, a former U.S. regulation official for the Securities and Exchange Commission, points to the power of AI to highlight key indicators of trafficking from hundreds of thousands of sources, providing better alerts and more likely real risks.

“For example, law enforcement can look at young women of a certain age entering the country from certain high-risk jurisdictions. Marry that up with social media and young people missing from home, or people associated with a false employment agency or who think they are getting a nanny job, and you start to develop a complete picture. And the information can be brought up all at once, rather than an analyst having to go through the Dark Web.”

To report suspected human trafficking to Federal law enforcement, call 1-866-347-2423.

To get help from the National Trafficking Hotline call 1-888-373-7888 or text HELP or INFO to BeFree (233733).

Learn more about AI-powered Radiance and its risk sensing capabilities for issues like human trafficking.

Solving the National Crisis of Veteran Suicide

Solving the National Crisis of Veteran Suicide

More than 6,000 veterans committed suicide in 2017 – an average of 17 suicides a day.

Veteran Suicide Rate is 1.5 times the rate of non-veterans

That number was recently reported in the 2019 National Veteran Suicide Prevention Annual Report, along with these equally sobering statistics: 

  • The veteran suicide rate is 1.5 times the rate for non-veterans.
  • Veterans ages 18-34 had the highest suicide rate (44.5 per 100,000). Overall, the suicide rate for this age group has increased by 76 percent since 2005.
  • In addition to the veteran suicides, there were 919 suicides among never federally activated National Guard and Reserve members, an average of 2.5 per day.

The report reinforces the magnitude of the crisis facing former members of our military. And, it comes just months after a renewed call for a comprehensive approach to address this national tragedy.

PREVENTS: A Comprehensive National Approach

In March 2019, the Trump Administration announced its FY 2020 budget proposal of $9.4 billion for veteran mental health services, including $222 million for suicide-prevention outreach, a $15.6 million increase over 2019.

That same month, President Trump also issued an Executive Order on a National Roadmap to Empower Veterans and End Suicide (PREVENTS). The PREVENTS Initiative calls for the development of a comprehensive public health strategy across all levels of government, and the private and non-profit sectors.

The goal is to understand the underlying factors of suicide, cultivate active engagement with veterans, and increase the timely identification of risk and intervention for those in need. 

Increasing Timely Identification and Intervention

A national research strategy is among the key components of PREVENTS. The Office of Science and Technology Policy (OSTP) is tasked with leading efforts to improve the coordination, monitoring, benchmarking, and execution of suicide-related data and research.

In its Request for Information, the OSTP announced its milestones and metrics would be focused on improving the ability to identify individual veterans and groups of veterans at greater risk of suicide and draw upon technology to capture and use health data from non-clinical settings to help target prevention and intervention strategies.

AI as an Early Detection System

Some experts believe that machine learning can be part of the solution when it comes to early intervention and risk prediction, suggesting that AI can be an early detection system by identifying and monitoring behaviors indicative of suicidal ideation.

One study they point to, conducted by researchers at the New York University School of Medicine, and funded by a grant from the U.S. Army Medical Research and Acquisition Activity, used speech-based algorithms to help detect posttraumatic stress disorders (PTSD) from warzone-exposed veterans.

Speech-based Algorithms Help Identify PTSD

The study analyzed audio recordings of clinical interviews, creating 40,526 speech features that were input into an algorithm, and ultimately shaved down to eighteen specific markers indicative of the potential for PTSD.  The algorithm correctly classified cases 89.1% of the time based on slower, more monotonous speech characteristics, less change in tonality, and less variation in activation.

Machine learning and AI are also being used to better analyze the Veterans Administration’s (VA) electronic health records to identify key factors related to suicide risks.

Deep Learning Neural Networks Predict Risk Based on Physicians’ Notes

A collaboration between the VA, the Department of Energy (DOE) and researchers at Lawrence Berkeley National Lab, focused on building deep learning neural networks that could distinguish between patients at high risk and those who are not based on physicians’ and discharge notes. 

Among the challenges was the noisy data sets that included structured data such as lab work and procedures and the unstructured data, like handwritten notes. But, as one researcher on the project pointed out, the value is in that unstructured data: 

“We believe that, for suicide prevention, the unstructured data will give us another side of the story that is extremely important for predicting risk — things like what the person is feeling, social isolation, homelessness, lack of sleep, pain, and incarceration. This kind of data is more complicated and heterogeneous, and we plan to apply what we have learned …to help VA doctors better decide who is at high risk and who they need to reach out to.”

The Path Ahead

The PREVENT Task Force’s mandate is to submit a proposed roadmap forward by next March.  The possibilities presented by AI and machine learning suggest that this technology should be a key area of focus, with continued investment and research.

While early identification of suicidal behaviors and risk is just one piece of helping end this national tragedy, it is a critical component of the overall strategy – and AI can play an important role.

To contact the Veteran Crisis Line, callers can dial 1-800-273-8255 and select option 1 for a VA staffer. Veterans, troops, or their family members can also text 838255 or visit for assistance.

The National Suicide Prevention Lifeline is 1-800-273-8255.

The Case for See Something, Say Something: 24 Hours at Lumina

The Case for See Something, Say Something: 24 Hours at Lumina

It was any given Tuesday afternoon at Lumina.

And then the S4 alert came. 

(S4 is a mobile app that allows people to report concerning behaviors in real time. It’s short for See Something, Say Something).

The first alert: Tuesday Afternoon 

This alert was from a high school student. 

The student expressed concern that a best friend was at risk for suicide. 

It turns out that the two students had recently lost another close friend to suicide.  Since that time, the friend at risk had been distant and negative, and showed other warning signs, which you can read more about from the American Foundation for Suicide Prevention.

The student who sent the S4 alert wanted to make sure the best friend got help before it was too late. 

Not surprisingly, the student wished to remain anonymous.  But the student shared the school information and the name of the friend.  Our S4 app also validated the location from which the alert was sent.

80% of those considering suicide give some sign of their intentions

This report was a serious concern.  Statistics show that 80% of those considering suicide give some sign of their intentions, and often those signs are communicated to the people closest to them.

We acted immediately, calling the school and sharing the information with the administration, who confirmed the recent suicide and thanked us for the report.

A person in time of crisis would get the help they need. 

The second alert:  Wednesday late morning

Just 20 hours later, we received another S4 alert.

This one was different. 

The person reporting the concern had innocently moved a postal package for a neighbor.  Then, the person noticed that the package had a marking indicating that it was from a company that sells bulletproof armor.

What to do with this information?  Buying armor isn’t illegal.  But, why was this neighbor concerned?  Was there something else to the report?

More than 75% of the attackers in mass violence events exhibit concerning behaviors

According to the U.S. Secret Service’s analysis of Mass Attacks in Public Spaces last year,  78% of the attackers in mass violence events exhibited behaviors that caused concern in others.

We decided to do more research. 

We ran the name of the subject the package was delivered to through our Radiance Open Source platform, OS-INT

Through a search of all the open source data on the Internet, OS-INT found publicly available social media images of the subject holding an IED, raising concerns that perhaps there was more to investigate.

We sent the S4 report, and the findings from OS-INT to the authorities, so they could determine appropriate next steps.

The third report:  Wednesday afternoon

While we were working the bulletproof armor S4, another alert came in.

Again, it was a report concerning potential mass violence.

But this time, it was at a school.

The report indicated that a student had discussed bringing a gun to school the next day. The report included the student’s name, and the school that he attended.

93% of the attackers in mass violence events made threatening or concerning communications

The same Secret Service report we mentioned previously, tells us that of the mass attacks in 2018, 93% of attackers made threatening or concerning communications prior to the act.

That is why – like the bulletproof armor S4 report – we ran the student’s name through Radiance OS-INT. 

Radiance quickly sent us a link to a publicly available YouTube channel where a person with the same name as the student shows himself executing a shooting rampage in a video game.  We thought this was important additional information to share with the local authorities.

Within minutes, we called the police, and learned they had received another tip surrounding the same student and were following up on the reports.

The power of S4

When we launched our S4 app, we knew the value it would bring to our clients as they work to keep their school and corporate campuses safe.

But we also understood the potential power it had for the broader public.

We knew we had to make the app available to others.  And it had to be free of charge.

Since making our app available, we have had thousands of downloads and hundreds of reports.

The truth is, we never like it when those reports come in. The thought that someone might want to do harm to themselves, or to others, keeps us up at night.

Be the light in your community

But, we know See Something, Say Something works.  And we’re committed to using our technology to help make a difference.

We encourage you to do the same.

Download the app today at the App Store or from Google Play, and help be the light in your community.

And, if you are in a suicide crisis, or know someone who is the National Suicide Prevention Lifeline is 1-800-273-8255.

The Increasing Threat to the Global Energy Supply

The Increasing Threat to the Global Energy Supply

This month’s attack on Saudi Arabia’s Abqaiq oil processing facility, which is the world’s largest and accounts for five percent of global oil supplies, resulted in one of the biggest oil price increases  ever recorded. 

More importantly, it demonstrated that the world’s energy infrastructure is vulnerable, can be severely disrupted and is an increasingly likely target for bad actors.

Recent Attacks Reinforce the Threat

Other recent examples – of both cyber and physical attacks – reinforce the threat.

In 2008, an alleged cyber attack blew up an oil pipeline in Turkey, shutting it down for three weeks.  In 2015, a Distributed Denial of Service (DDos)  attack brought down a section of the Ukrainian power grid — for just six hours, but substations on the grid had to be operated manually for months.  Another attack in the Ukraine occurred just a year later, reportedly carried out by Russian actors. And, the Abqaiq facility itself had been the target of a thwarted Al Qaeda suicide bomber attack in 2006.

Threats to Physical Security

A 2018 report by the United Nations Office of Counter-Terrorism outlined the most intuitive physical threats to critical infrastructure, including the energy sector, involved the use of explosives or incendiary devices, rockets, MANPADs, grenades and tools to induce arson.

That same report noted that the energy sector has witnessed sustained terrorist activity through attacks perpetrated by Al Qaeda and its affiliates on oil companies’ facilities and personnel in Algeria, Iraq, Kuwait, Pakistan, Saudi Arabia and Yemen.

Increasing Intensity of DDoS Attacks

In addition to physical threats, it is estimated that by 2020, at least five countries will see foreign hackers take all or part of their national energy grid offline through Permanent Denial of Service (PDoS) attacks. And, DDoS attacks like those in the Ukraine are becoming increasingly severe.  Studies show that the number of total DDoS attacks decreased by 18 percent year-over-year in Q2 2017.  At the same time, there was a 19 percent increase in the average number of attacks per target.

U.S. is the “Holy Grail”

Disruption of the U.S. power grid is considered the “holy grail,” and experts predict that the energy industry could be an early battleground, not only the power sector, but the nation’s pipelines and the entirety of the supply chain. 

In fact, last year the Department of Homeland Security (DHS) and the Federal Bureau of Investigation (FBI) publicly accused the Russians of cyberattacks on small utility companies in the United States.  In a joint Technical Alert  (TA), the agencies said Russian hackers conducted spear phishing attacks and staged malware in the control rooms with the goal of gathering data to create detrimental harm to critical U.S. infrastructure.

900 “Vulnerabilities” Found in the U.S. Energy Systems

This specific incident aside, DHS’s Industrial Control System Computer Emergency Response Team found nearly 900 cyber security vulnerabilities in U.S. energy control systems between 2011 and 2015, more than any other industry.  It’s not surprising that the international oil sector alone is increased investments on cyber defenses by $1.9 billion in 2018. 

Investment in Physical Security Will Reach $920 billion

With any disruption to the global or national energy supply having serious implications for virtually all industries, especially critical ones like healthcare, transportation, security, and financial services, one report projects that the global critical infrastructure protection market will be worth $118 billion by 2028.

Physical security is expected to account for the highest proportion of spending, and cumulatively will account for $920 billion in investment.

Artificial Intelligence: A Security “Pathway” for the Future

Experts suggest that these investments should include next generation technologies for both physical and cyber security purposes. As one expert put it: “Automation, including via artificial intelligence, is an emerging and future cyber security pathway.”

In addition to the role that automation, artificial intelligence and machine learning can bring to identifying and predicting a physical or cyber attack, research shows that it can also help manage the rising costs associated with it. A study found that only 38 percent of companies are investing in this technology – even though after initial investments, it could represent net saving of $2.09 million.

Learn more about AI-driven Radiance and how it can help identify and predict physical and cyber threats to the energy infrastructure.