Memorandum Updated January 21, 2003

SUBJECT: Funding for Total Information Awareness programs

FROM: Amy Belasco

Consultant on the Defense Budget

Foreign Affairs, Defense, and Trade Division

Controversy about Level of Funding

In recent months, there has been some disagreement about the funding for DARPA’s Information Awareness Office and for the Total Information Awareness system. In a press conference on November 20, 2002, Under Secretary of Defense for Acquisition, Logistics, and Technology Edward C. (Pete) Aldridge, stated that funding for the Total Information Awareness System (TIA) is $10 million in FY2003.¹ The Electronic Privacy Information Center, a non-profit organization specializing in privacy issues, calculated the “total cost of TIA,” in FY2001-2003 as $245 million.² Recent press articles also cited funding of over $200 million over three years.³

These differences reflect confusion between funding for the DARPA R&D specifically labeled “Total Information Awareness, “ which is dedicated to integrating various R&D efforts to produce a prototype Total Information Awareness system, and the funding for the various R&D technology efforts managed by the Information Awareness Office (IAO) that could become part of that system.

Current Funding for Programs Managed by the Information Awareness Office and for the Total Information Awareness System

Established in January 2002 under retired Admiral John Poindexter, the mission of the Information Awareness Office is to develop new tools to detect, anticipate, train for, and

provide warnings about potential terrorist attacks.⁴ The objective is to integrate these tools into a prototype Total Information Awareness system or systems within three to five years to provide better intelligence support to senior government officials. If proven effective, DOD Under Secretary Aldridge suggested that the TIA technology prototypes will be turned over to “intelligence, counterintelligence and law enforcement communities as a tool to help them in their battle against domestic terrorism.”⁵

Funding for DARPA, as for the Research, Development, Test & Evaluation (RDT&E) programs for the services, is authorized and appropriated annually at the account level. In the case of DARPA, funding is included within RDT&E, Defensewide account.⁶ The TIA system, like other R&D efforts, is not specifically identified in statutory language. Congressional intent about the funding levels for individual R&D efforts, however, may be included in committee conference reports. The FY2003 DOD authorization and appropriation conference reports did not include any specific language about the TIA system.⁷

FY2001-FY2003 Funding Levels According to DARPA, technology developed in some or all of the sixteen R&D efforts managed by the Information Awareness Office may be integrated into the Total Information Awareness system.⁸ DARPA’s FY2003 request for the R&D efforts managed by the Information Awareness Office totaled $137.5 million in FY2003 (see Table 1 below), including $10 million for the integrative efforts specifically labeled the Total Information Awareness System.

DARPA’s budget materials state that TIA will integrate technology and components from 9 of the 16 R&D efforts (including the integration itself) managed by the Information Awareness Office. According to DARPA, TIA is “the assured transition of a system-level prototype that integrates technology and components developed in other DARPA programs including Genoa and Genoa II . . . TIDES . . ., Genisys, EELD, WAE, HID, and Bio-Surveilance . . . “⁹ (See Table 2 and Appendix for funding and description of these R&D efforts). Funding for the 9 R&D efforts totals $110.6 million in FY2003, $83.8 million in FY2002, and $65.0 million in FY 2001 (see Table 1). (Three follow-on machine translation efforts under the Information Awareness Office will probably also be incorporated into the TIA system but funding for those programs is not identified separately in 2003.) According to DARPA, the TIA system may also exploit the results of other R&D efforts that are under the Information Awareness office, other DARPA efforts, or R&D conducted outside of DARPA.¹⁰

Table 1. Funding for Information Awareness Office and for Total Information Awareness Technology, FY2001-FY2003^a

In millions of dollars

Total Funding	FY 2001	FY2002	FY2003
Total Information Awareness System integrationb	0.0	0.0	10.0
Technology Programs Supporting TIA c	65.0	83.8	110.6
Information Awareness Office programs	80.0	99.5	137.5

Sources and Notes:

See DARPA, RDT&E Descriptive Summaries for FY2003 (or the R-2), available at web site, http://www.dtic.mil/comptroller/fy2003budget/budget_justification/pdfs/ rdtande/darpa_vol1.pdf

^a FY2003 level reflects DARPA’s request.

^b TIA is shown by DARPA as a specific R&D effort in Project CCC-01 in Program Element 603760E.

c Includes the 9 R&D efforts identified in DARPA’s FY2003 budget justification materials as linked to the TIA system including four data mining efforts (Human Identification at a Distance, Evident Extraction and Link Discovery, Genisys, Bio-surveilance), machine translation of languages (TIDES), and four decisionmaking tools (Wargaming the Asymmetric Environment, Project Genoa, Genoa II, and Total Information Awareness); see appendix for description of these efforts.

Although the TIA system was first proposed as an integrated entity in the FY2003 budget shortly after establishment of the Information Awareness Office, some of the R&D efforts that may become part of that system have been underway for a number of years. In fact, several of the R& D efforts, e.g. Project Genoa and machine translation of languages, were funded as early as FY 1997. For comparative purposes, Table 1 above and the more detailed Table 2 below shows funding from FY2001 through FY2003 for all the elements now managed by the Information Awareness Office that could become part of the Total Information Awareness system as well as those definitely tied to the TIA system.

DARPA’s FY2003 request for projects of the Information Awareness Office of $137.5 million compares to a funding level of $99.5 million in FY2002 and $80 million in FY2001. Funding for the three-year period, FY2001 - FY 2003, for the various components that could be integrated into the TIA system totaled $317.0 million. The increase in FY 2003 primarily reflects new starts in FY2003 for Genisys, a comprehensive data mining effort, MIDGET, a system designed to protect against misinformation that is deliberately inserted into open databases, Rapid Analytic Wargaming, a tool for decision makers, and the TIA integration effort (see Table 2 below and Appendix).

Final FY2003 Funding Level Still Unknown The final appropriated level for R&D efforts managed by the Information Awareness Office FY2003 could be slightly lower than the $137.5 million requested because Congress cut certain DARPA programs that could affect those efforts. DARPA’s FY2004 budget proposal to Congress will show how those cuts have been allocated, and whether the funding for TIA components has been reduced from the level requested in FY2003.¹¹

FY 2001 - FY 2003 Funding for Individual R&D Efforts Based on their primary purpose, CRS grouped the sixteen R&D efforts managed by the Information Awareness Office into the four categories below. Table 2 below shows the funding for FY2001-FY2003 for the individual R&D efforts managed by the Information Awareness Office, for those 9 systems that are to be integrated into the TIA system, and for the four categories.¹² The appendix briefly describes each of the sixteen individual R&D efforts.

New data mining technologies These R&D efforts are designed to develop technologies that would be capable of sifting through large data bases to detect patterns associated with terrorists’ activities. Total funding for these efforts was $29.2 million in FY2001, $38.2 million in FY 2003 and $53.0 million in FY 2003. The increases in FY 2002 and FY2003 reflect the initiation of the Bio-surveillance effort in FY 2002 and the Genisys program in FY 2003.

New machine translation technologies These R&D efforts are intended to develop new software technology to translate large volumes of foreign language material, both written and oral, that would be collected from sources ranging from electronic sources to battlefield transmissions. At $36 million annually, funding for these efforts was stable between FY 2001 and FY2003.

Protection of Critical Information Infrastructure These R&D efforts are intended to protect DOD’s information infrastructure and detect mis-information in open-source data that DOD may collect. Funding in this area grew from zero in FY 2001 to $2.0 million in FY 2002 with the initiation of DefenseNet, and jumped to $9.5 million with the addition of the Mis-Information Detection and Generation effort.

Tools for High-Level Decision Makers These R&D efforts are designed to develop various tools from war-gaming simulations to collaborative decision making tools that can be used by high-level decision makers to anticipate, train for, pre-empt, or react to terrorist acts. Funding for these efforts increased from $14.4 million to $23.5 million in FY2002 with the doubling in the funding level for Wargaming the Asymmetric Environment, and increased further to $39.5 million with the initiation of the integrative effort, the Total Information Awareness System.

Future Funding Levels for Information Awareness Office Programs

Although DARPA does not specifically identify future funding levels for the R&D efforts that are managed by the Information Awareness Office, future funding levels can be estimated from DARPA’s FY2003 budget submissions. Such estimates could, of course, change based on changes in emphasis by DARPA or the success or failure of individual technology efforts.

To estimate future funding levels, CRS relied on patterns in funding for TIA-related R&D between FY 2001 and FY2003. Based on those trends, annual funding for TIA-related R&D between FY 2004 to FY 2007 may average about $145 million, a level close to that of FY 2003.¹³ That level would follow the relatively stable pattern of many DARPA technology efforts. If that trend were to hold, DARPA would spend about $575 million for TIA-related R&D in the next four years in addition to the $317 million spent from FY2001-FY2003.Additional information about future funding levels for these R&D efforts will be provided when DARPA’s FY2004 budget is submitted to Congress.¹⁴

Table 2. FY2001-FY2003 Funding for Information Awareness Office and Technology Supporting Total Information Awareness System^*

* identifies R&D linked specifically by DARPA to TIA System

in millions

Major Purpose by Category	Project	Program Element	FY2001	FY2002	FY2003 Request
Data Mining Technologies Subtotal:			29.2	38.2	53.0
Human ID at a Distance*	Asymmetric Threat ST-28	602301E	11.8	15.9	14.5
Evidence Extraction and Link Discovery (EELD)*	ST-28	602301E	17.3	14.4	14.0
Genisys*	ST-28	602301E	0.0	0.0	11.0
Bio-surveillance*	ST-28	602301E	0.0	8.0	13.5
Machine Translation of Languages Subtotal:			36.5	35.8	35.5
Translingual Information Detection, Extraction and Summarization (TIDES)* and Effective, Affordable, Reusable Speech-to-Text (EARS), and Multispeaker Environments (MUSE) and Global Autonomous Language Exploitation (GALES)a	Intelligent Systems and Software, ST-11	602301E	21.5	22.1	22.1
Babylon and Communicatora	ST-11	602301E	15.0	13.7	13.4
Protection of DOD’s Information Infrastructure Subtotal:			0.0	2.0	9.5
DefenseNet (DNET)b	ST-28/ST-11c	602301E	0.0	2.0	3.0
Mis-Information Detection and Generation (MIDGET)	ST-28	602301E	0.0	0.0	6.5
Tools for High-Level Decision makers Subtotal:			14.4	23.5	39.5
Rapid Analytic Wargaming (RAW)	ST-11	602301E	0.0	0.0	4.0
Wargaming the Asymmetric Environment (WAE)*	Command & Control Info. Systems, CCC-01	603760E	6.9	15.8	18.5
Project Genoa/Genoa II c*	CCC-01	603760E	7.5	7.6	7.0
Total Information Awareness System d*	CCC-01	603760E	0.0	0.0	10.0
Technology Supporting TIA System*	NA	NA	65.0	83.8	110.6
Three-Year Total, FY 2001-FY2003*	259.4
Information Awareness Office Total	NA	NA	80.0	99.5	137.5
Three-Year Total, FY2001-FY2003:	317.0

Sources and Notes :

DARPA, and home page for Total Information Awareness Office program; www.defenselink.darpa.mil/iao/programs.

See DARPA, RDT&E Descriptive Summaries for FY2003 (or the R-2), available at web site, http://www.dtic.mil/comptroller/fy2003budget/budget_justification/pdfs/ rdtande/darpa_vol1.pdf

^aFunding for individual components not shown in DARPA’s FY2003 budget justification.

^b DefenseNet transfers from Project ST-28 in FY2002 to Project ST-11 in 2003; see DARPA\'s R-2, p. 90; or,

http://www.dtic.mil/comptroller/fy2003budget/budget_justification/pdfs/rdtande/darpa_vol1.pdf

c Funding for GenoaII starts in FY2003.

^dTotal Information Awareness is the integrative effort.

Appendix: Description of R&D Efforts* Managed by the Information Awareness Office

By Category

(* = R&D efforts specifically linked to the TIA system by DARPA)

Data Mining Technologies

• Human Identification at a Distance (HumanID).* This project aims to use information from sensors about human characteristics such as gait or face, to identify individuals at any time of the day or night and in all weather conditions, for instance, within a large crowd.

• Evidence Extraction and Link Discovery (EELD).* This project is an effort to identify terrorist groups by developing a suite of technologies to detect patterns between people, organizations, places and things from intelligence messages and law enforcement records , and then use those patterns or links to gather additional information from vast amounts of textual or transactional data including web sites, sensor data, and news reports.

• Genisys.* This project is a new effort in 2003 to put together old and new databases so that they can be readily queried. This “ultra-large all-source information repository” could include information about potential terrorists and possible supporters, purchase of terrorist types of material, training and rehearsal activities, potential targets, and status of defenses, as well as research into methods of protecting privacy.¹⁵

• Bio-surveillance.* This project is an effort to collect and analyze information from non-traditional human, agricultural and animal health data bases in order to develop indicators and models, and set up a prototype bio-surveillance system for a citywide area like Norfolk, Virginia in order to increase DOD’s ability to detect a clandestine biological warfare attack.

Machine Translation Projects

• Translingual Information Detection, Extraction and Summarization (TIDES).* TIDES is designed to get critical information quickly for intelligence analysts and operators by developing tools that can rapidly find, summarize, and translate key information in foreign languages.

• Effective Affordable Reusable Speech-to-Text (EARS): Anticipated to increase the speed of translation from oral sources by ten to 100-fold (including broadcasts and telephone), as well as extract clues about the identity of speakers, EARS is intended to serve the military, intelligence and law enforcement communities.

• Multispeaker Environments (MUSE) and Global Autonomous Language Exploitation (GALE): MUSE and GALE are successor programs to EARS, MUSE is to produce transcripts from command centers and meeting rooms and GALE is to develop techniques for detecting key intelligence in massive amounts of foreign language transmissions.

• Communicator: Designed as a tool for military personnel in the field to get logistical support and tactical information, prototypes of this “smart phone” that was originally developed for the Marine Corps have already been deployed on Navy ships.

• Babylon: Another battlefield system likely to be deployed in Afghanistan in the next few months, Babylon is intended to aid those in the field by translating foreign phrases for the service member.¹⁶

Protection of Critical Information Infrastructure

• DefenseNet (DNET): This effort is intended to increase the security and performance of DOD’s information infrastructure in order to handle large volumes of information.

• Mis-Information Detection and Generation (MIDGET): A new project in 2003, this effort is designed to detect and hence reduce DOD’s vulnerability to mis-information about adversaries that appears in open-source data.

Tools for High-Level Decision Makers

• Rapid Analytic Wargaming (RAW): This project is intended to develop gaming technologies that simulate asymmetric threats to be used by the major commands in training and operational settings.

• War Gaming the Asymmetric Environment (WAE).* This effort is an initiative to develop tools and models to help analysts and decision makers predict and anticipate the behavior and the reactions of terrorists to U.S. actions.

• GENOA/GENOA II.* Project Genoa was designed to improve collaborative reasoning, estimate plausible futures, and create actionable options among intelligence analysts in various organizations. Genoa II is intended to improve the collaboration between people and machines in order to improve information support that is provided by intelligence analysts to policy makers at the military command level, high level DOD civilian officials, NSA and the Joint Chiefs of Staff, particularly for dealing with terrorist threats.

• Total Information Awareness.* TIA is to integrate some or all of the efforts above into a prototype system or a series of prototype systems that will create and exploit large-scale, counter-terrorist data bases, develop new analytical tools for mining those data bases and create new models and analytical techniques to improve our ability to detect, anticipate, pre-empt, and respond to terrorist attacks. R&D efforts specifically linked to the TIA system in FY2003 are Human ID at a Distance, EELD, Genisys, Bio-surveillance, TIDES, WAE, Project Genoa and Genoa II, and the TIA integrative effort.