Microsoft’s planned East Africa cloud region in Kenya, announced in May 2024 with G42 as part of a $1 billion digital investment package, has reportedly been delayed after talks over guaranteed government payments and power requirements failed to produce terms acceptable to all sides. The episode is not just another stalled construction story. It exposes the harder bargain beneath the AI infrastructure boom: hyperscale cloud regions are being sold as national development engines, but they increasingly arrive with the financial expectations of utility-scale industrial projects.
When Microsoft and G42 unveiled the Kenya plan in 2024, the pitch was almost perfectly tuned to the moment. It promised an East Africa Azure region, a geothermal-powered data center campus in Olkaria, local-language AI work, digital skills programs, connectivity investment, and a government partnership around secure cloud services. For Kenya, it was framed as a leap into the cloud economy; for Microsoft, it was another marker in the company’s campaign to turn AI demand into physical global reach.
That framing mattered because this was not merely a co-location facility with a Microsoft logo on the door. A cloud region is a promise of local capacity, lower latency, data residency options, and enterprise-grade services close enough to customers to change procurement calculations. In markets where government agencies, banks, telecoms, hospitals, and large businesses worry about sovereignty and reliability, a local region can become a catalyst.
But the project also carried all the burdens of the modern data center economy. These campuses are power-hungry, capital-intensive, and dependent on long-term demand assumptions that must survive politics, currency pressure, procurement cycles, and grid constraints. The Bloomberg report, carried by Reuters, suggests that the apparently inspirational part of the plan — building cloud infrastructure in a fast-growing African market — collided with the less photogenic question of who would guarantee enough revenue to justify it.
That is the tension at the center of the story. Microsoft’s public cloud sells elasticity to customers, but the infrastructure underneath is anything but elastic. Servers, substations, fiber, cooling systems, land, and energy contracts require commitments long before developers and government ministries begin spinning up workloads.
That request is not inherently unreasonable. Governments routinely underpin infrastructure through offtake agreements, minimum revenue guarantees, power purchase agreements, availability payments, and long-term service contracts. Airports, roads, power plants, rail systems, and broadband networks often become financeable because a state or state-backed entity absorbs part of the demand risk.
But cloud infrastructure is politically different. When a government guarantees payment for unused road capacity, citizens may still see a physical road. When it guarantees cloud capacity, the benefit is more abstract, and the downside can look like a blank check to a foreign technology giant. In a country managing competing budget priorities, a multiyear public commitment to consume cloud services at a level set by hyperscale investors is a hard sell.
That appears to be where the Kenya project ran into trouble. According to the report, talks broke down when the government could not provide guarantees at the level Microsoft requested. Kenya’s principal secretary at the Ministry of Information, John Tanui, was quoted as saying the project had not failed or been withdrawn, while also acknowledging that the intended scale still required structuring and that power requirements remained under discussion.
That distinction is important. This is not, at least on the reported record, a cancellation. It is a renegotiation under pressure. The difference matters for investors, customers, and regional IT leaders who were watching the project as a signal that East Africa might soon join the map of first-class cloud locations rather than remain dependent on more distant regions.
G42 is not just another data center developer. The Abu Dhabi-based AI company became a strategic Microsoft partner after Microsoft announced a major investment in it in 2024, a deal widely understood as part commercial expansion and part geopolitical alignment. The United States, the UAE, and major technology firms have all been navigating a world in which advanced AI infrastructure is treated as strategically sensitive.
That makes the Kenya project more than a local buildout. It is part of a broader pattern in which AI infrastructure is being exported through alliances: American cloud platforms, Gulf capital, African energy resources, and national governments seeking digital transformation. Everyone arrives with a different objective, and not all of those objectives price risk the same way.
Microsoft wants Azure consumption and AI relevance. G42 wants to expand its footprint and credibility beyond the Gulf. Kenya wants infrastructure, jobs, skills, investment, and a stronger position as a regional digital hub. Customers want reliable services at commercially viable prices. The friction begins when those ambitions must be translated into contracts.
The reported demand for guaranteed payments is the point where rhetoric becomes finance. A government may welcome a $1 billion initiative, but that does not mean it can or should underwrite hyperscale demand forecasts. The cloud industry’s growth story has been built on the idea that demand will arrive. In emerging markets, investors increasingly want proof before they pour concrete.
Yet renewable does not mean simple. A data center does not merely need a green megawatt-hour on paper. It needs reliable capacity, transmission, redundancy, cooling, water strategy, permitting, grid interconnection, and a commercial structure that can survive years of operation. The original plan reportedly contemplated significant initial capacity, and that scale changes the conversation from “can Kenya produce geothermal energy?” to “can the project secure the right power at the right reliability and price?”
Tanui’s reported comment that power requirements remain under discussion is therefore not a footnote. It is one of the central constraints. AI-era data centers are forcing governments and utilities to make choices about where scarce firm power goes, how quickly grids can be upgraded, and whether industrial customers should receive preferential arrangements.
This is where the clean-energy story can become politically awkward. A data center powered by geothermal energy sounds like a win for sustainability, but citizens and businesses will still ask whether the same energy could support manufacturing, households, hospitals, or local enterprises. If public guarantees are also on the table, the question becomes sharper: is the state enabling a national digital platform, or subsidizing an export-oriented cloud asset?
Microsoft is hardly alone in facing that scrutiny. The entire hyperscale industry is now running into power as a gating factor. In Northern Virginia, Dublin, Singapore, the Netherlands, and parts of the United States, data center growth has triggered debates over grid capacity, water use, land, and public benefit. Kenya’s version of the debate has its own local politics, but the underlying pattern is global.
For WindowsForum readers, the Azure region angle is the most concrete part of the story. A local region can affect where Microsoft 365, Dynamics, Azure Virtual Desktop, backup, disaster recovery, identity, analytics, and AI services are deployed. It can also influence procurement rules for regulated industries that need clearer data residency and service continuity assurances.
If the East Africa region is delayed, that does not stop Kenyan or regional organizations from using Azure. It does, however, preserve the disadvantages of distance. Latency-sensitive workloads remain harder to justify. Data protection conversations remain more complicated. Disaster recovery architectures may continue to depend on regions outside the immediate market.
The delay also matters symbolically. Cloud regions are signals. When Microsoft, Amazon, or Google places a region in a market, it tells customers that the provider sees enough demand, stability, and infrastructure depth to make a long-term bet. A scaled-back project would not erase Kenya’s digital ambitions, but it would make the original 2024 announcement look more like aspiration than execution.
That distinction matters because governments increasingly use cloud announcements as proof of economic momentum. Ribbon-cutting promises travel faster than substations. The hard part comes later, when the parties must align capacity, price, procurement, sovereignty, and public accountability.
AI has accelerated the shift. The demand for compute is enormous, but it is also uncertain. Enterprises are experimenting aggressively, but not every proof of concept becomes sustained consumption. Governments want sovereign AI capabilities, but public budgets do not move at the speed of GPU procurement. Data centers must be financed before the demand curve is fully proven.
In mature markets, hyperscalers can lean on existing enterprise bases, deep capital markets, and large ecosystems of partners. In emerging markets, the first major cloud region may need an anchor customer strong enough to convince financiers that utilization will materialize. Often, that anchor is the public sector.
That is why guaranteed capacity payments are so revealing. They are a mechanism for turning expected demand into bankable revenue. They also move risk from the provider and its partners toward the state. Whether that is sensible depends on the price, the services covered, the flexibility of the contract, the public benefit, and the transparency of the arrangement.
If the reported guarantee request was too large for Kenya, the issue may not be whether the country wants cloud infrastructure. It may be whether the proposed scale arrived ahead of demand that could be credibly contracted. Scaling back, if it happens, would be a rational response to that mismatch rather than a humiliation.
The Kenya report shows the limits of that model. A global company can announce a standardized ambition, but it cannot standardize every country’s fiscal capacity, regulatory environment, power market, or political tolerance for guarantees. The more Microsoft frames cloud regions as national development infrastructure, the more it must deal with the messy democratic and fiscal questions that surround national infrastructure.
There is also a reputational risk. Microsoft’s cloud credibility rests partly on trust: trust that announced regions will arrive, trust that services will be durable, trust that governments and enterprises can plan around the roadmap. If a high-profile project slips, customers will ask whether other emerging-market announcements should be treated as firm commitments or strategic signaling.
That does not mean Microsoft should avoid ambitious markets. Quite the opposite. The next phase of cloud growth will depend on reaching countries and regions underserved by current infrastructure. But the company may need to be more explicit about the difference between a signed intention, a financed build, a contracted cloud region, and an operational launch.
For IT buyers, that distinction is practical. Roadmaps influence architecture. If a bank expects a local Azure region within two years, it may design modernization plans differently. If that region becomes uncertain, the bank needs a fallback strategy. Ambiguous announcements can create real planning costs.
The correct answer may be a smaller first phase, a broader base of anchor tenants, or a staged capacity model that ties expansion to actual demand. Microsoft and G42 may prefer a larger commitment because scale improves economics, but a smaller, credible deployment could be better than a grander project that cannot clear the financing and procurement hurdles.
Kenya also has leverage. Its geothermal resources, regional position, talent base, and role as an East African technology hub are real assets. Microsoft and G42 chose Kenya for reasons that have not vanished because negotiations became difficult. If anything, the continued talks suggest both sides still see value in finding a workable structure.
The government’s challenge is to avoid confusing prestige with policy. A local Azure region would be valuable, but it is not valuable at any price. The public interest depends on whether the project supports domestic capability, improves service access, strengthens resilience, and leaves Kenya with infrastructure and skills that outlast the announcement cycle.
That means transparency matters. Citizens do not need every commercial detail, but they deserve clarity about public obligations, power allocation, procurement commitments, and expected benefits. The more cloud infrastructure resembles critical national infrastructure, the less plausible it is to treat its financing as a purely private matter.
The most important lesson is that cloud geography remains strategic. Region selection affects latency, compliance, backup design, cost, identity architecture, and service availability. Organizations that depend on future regions should build plans that can survive schedule slips.
The second lesson is that AI demand is reshaping infrastructure priorities. Microsoft’s cloud roadmaps are increasingly tied to power availability and GPU-scale economics. That may affect not only where new regions appear, but also which services launch in those regions, how capacity is rationed, and how quickly advanced AI features become available outside the largest markets.
The third lesson is that government cloud strategy is now inseparable from energy strategy. Countries that want local AI infrastructure need reliable power, credible regulation, procurement frameworks, and enough demand aggregation to make projects financeable. Cloud adoption policy alone is no longer enough.
For enterprise customers in Kenya and neighboring markets, the prudent approach is to treat the East Africa region as promising but not guaranteed until Microsoft provides a firmer update. Design for portability where possible. Keep data residency assumptions explicit. Avoid building compliance narratives around infrastructure that is not yet operational.
Source: WKZO Microsoft’s African data center falters on payment demands, Bloomberg News reports
Microsoft’s Kenya Bet Was Always Bigger Than a Data Center
When Microsoft and G42 unveiled the Kenya plan in 2024, the pitch was almost perfectly tuned to the moment. It promised an East Africa Azure region, a geothermal-powered data center campus in Olkaria, local-language AI work, digital skills programs, connectivity investment, and a government partnership around secure cloud services. For Kenya, it was framed as a leap into the cloud economy; for Microsoft, it was another marker in the company’s campaign to turn AI demand into physical global reach.That framing mattered because this was not merely a co-location facility with a Microsoft logo on the door. A cloud region is a promise of local capacity, lower latency, data residency options, and enterprise-grade services close enough to customers to change procurement calculations. In markets where government agencies, banks, telecoms, hospitals, and large businesses worry about sovereignty and reliability, a local region can become a catalyst.
But the project also carried all the burdens of the modern data center economy. These campuses are power-hungry, capital-intensive, and dependent on long-term demand assumptions that must survive politics, currency pressure, procurement cycles, and grid constraints. The Bloomberg report, carried by Reuters, suggests that the apparently inspirational part of the plan — building cloud infrastructure in a fast-growing African market — collided with the less photogenic question of who would guarantee enough revenue to justify it.
That is the tension at the center of the story. Microsoft’s public cloud sells elasticity to customers, but the infrastructure underneath is anything but elastic. Servers, substations, fiber, cooling systems, land, and energy contracts require commitments long before developers and government ministries begin spinning up workloads.
The Payment Guarantee Is the Real Headline
The reported disagreement over guaranteed annual capacity payments goes to the heart of cloud economics. A hyperscale data center can be marketed as a national asset, but its balance sheet still needs predictable utilization. If Microsoft and G42 wanted the Kenyan government to commit to paying for a certain amount of capacity each year, they were effectively asking the public sector to become an anchor tenant.That request is not inherently unreasonable. Governments routinely underpin infrastructure through offtake agreements, minimum revenue guarantees, power purchase agreements, availability payments, and long-term service contracts. Airports, roads, power plants, rail systems, and broadband networks often become financeable because a state or state-backed entity absorbs part of the demand risk.
But cloud infrastructure is politically different. When a government guarantees payment for unused road capacity, citizens may still see a physical road. When it guarantees cloud capacity, the benefit is more abstract, and the downside can look like a blank check to a foreign technology giant. In a country managing competing budget priorities, a multiyear public commitment to consume cloud services at a level set by hyperscale investors is a hard sell.
That appears to be where the Kenya project ran into trouble. According to the report, talks broke down when the government could not provide guarantees at the level Microsoft requested. Kenya’s principal secretary at the Ministry of Information, John Tanui, was quoted as saying the project had not failed or been withdrawn, while also acknowledging that the intended scale still required structuring and that power requirements remained under discussion.
That distinction is important. This is not, at least on the reported record, a cancellation. It is a renegotiation under pressure. The difference matters for investors, customers, and regional IT leaders who were watching the project as a signal that East Africa might soon join the map of first-class cloud locations rather than remain dependent on more distant regions.
AI Has Made Cloud Regions More Political
Five years ago, a new cloud region was mostly discussed in terms of latency, compliance, and enterprise modernization. In 2026, it is also about AI capacity, geopolitical alignment, energy strategy, and industrial policy. Microsoft’s partnership with G42 sits directly inside that larger shift.G42 is not just another data center developer. The Abu Dhabi-based AI company became a strategic Microsoft partner after Microsoft announced a major investment in it in 2024, a deal widely understood as part commercial expansion and part geopolitical alignment. The United States, the UAE, and major technology firms have all been navigating a world in which advanced AI infrastructure is treated as strategically sensitive.
That makes the Kenya project more than a local buildout. It is part of a broader pattern in which AI infrastructure is being exported through alliances: American cloud platforms, Gulf capital, African energy resources, and national governments seeking digital transformation. Everyone arrives with a different objective, and not all of those objectives price risk the same way.
Microsoft wants Azure consumption and AI relevance. G42 wants to expand its footprint and credibility beyond the Gulf. Kenya wants infrastructure, jobs, skills, investment, and a stronger position as a regional digital hub. Customers want reliable services at commercially viable prices. The friction begins when those ambitions must be translated into contracts.
The reported demand for guaranteed payments is the point where rhetoric becomes finance. A government may welcome a $1 billion initiative, but that does not mean it can or should underwrite hyperscale demand forecasts. The cloud industry’s growth story has been built on the idea that demand will arrive. In emerging markets, investors increasingly want proof before they pour concrete.
Geothermal Power Was the Selling Point, Not a Magic Wand
The Kenya plan’s most compelling technical feature was its reliance on geothermal power. Olkaria is associated with Kenya’s geothermal resources, and the promise of a data center running entirely on renewable geothermal energy gave the project a cleaner narrative than the diesel-backed, coal-adjacent, or grid-stressed campuses often criticized elsewhere. For Microsoft, which has aggressive climate commitments and increasingly visible power needs, geothermal baseload is especially attractive.Yet renewable does not mean simple. A data center does not merely need a green megawatt-hour on paper. It needs reliable capacity, transmission, redundancy, cooling, water strategy, permitting, grid interconnection, and a commercial structure that can survive years of operation. The original plan reportedly contemplated significant initial capacity, and that scale changes the conversation from “can Kenya produce geothermal energy?” to “can the project secure the right power at the right reliability and price?”
Tanui’s reported comment that power requirements remain under discussion is therefore not a footnote. It is one of the central constraints. AI-era data centers are forcing governments and utilities to make choices about where scarce firm power goes, how quickly grids can be upgraded, and whether industrial customers should receive preferential arrangements.
This is where the clean-energy story can become politically awkward. A data center powered by geothermal energy sounds like a win for sustainability, but citizens and businesses will still ask whether the same energy could support manufacturing, households, hospitals, or local enterprises. If public guarantees are also on the table, the question becomes sharper: is the state enabling a national digital platform, or subsidizing an export-oriented cloud asset?
Microsoft is hardly alone in facing that scrutiny. The entire hyperscale industry is now running into power as a gating factor. In Northern Virginia, Dublin, Singapore, the Netherlands, and parts of the United States, data center growth has triggered debates over grid capacity, water use, land, and public benefit. Kenya’s version of the debate has its own local politics, but the underlying pattern is global.
East Africa Still Needs Local Cloud Capacity
None of this means the project is misguided. East Africa has a strong case for local cloud infrastructure. Businesses want lower latency. Governments want more control over sensitive data. Developers want regional services that do not force every workload through Europe, South Africa, or the Middle East. Banks, insurers, telecom operators, universities, and public agencies all benefit when advanced cloud services are closer to users.For WindowsForum readers, the Azure region angle is the most concrete part of the story. A local region can affect where Microsoft 365, Dynamics, Azure Virtual Desktop, backup, disaster recovery, identity, analytics, and AI services are deployed. It can also influence procurement rules for regulated industries that need clearer data residency and service continuity assurances.
If the East Africa region is delayed, that does not stop Kenyan or regional organizations from using Azure. It does, however, preserve the disadvantages of distance. Latency-sensitive workloads remain harder to justify. Data protection conversations remain more complicated. Disaster recovery architectures may continue to depend on regions outside the immediate market.
The delay also matters symbolically. Cloud regions are signals. When Microsoft, Amazon, or Google places a region in a market, it tells customers that the provider sees enough demand, stability, and infrastructure depth to make a long-term bet. A scaled-back project would not erase Kenya’s digital ambitions, but it would make the original 2024 announcement look more like aspiration than execution.
That distinction matters because governments increasingly use cloud announcements as proof of economic momentum. Ribbon-cutting promises travel faster than substations. The hard part comes later, when the parties must align capacity, price, procurement, sovereignty, and public accountability.
The Cloud Boom Is Learning the Language of Infrastructure Finance
The uncomfortable lesson from the reported Kenya dispute is that cloud expansion is beginning to look less like software rollout and more like infrastructure finance. That is a major change in how the industry presents itself. Cloud companies prefer to sell services, platforms, and innovation; data center developers and financiers talk about offtake, power, capacity, and risk allocation.AI has accelerated the shift. The demand for compute is enormous, but it is also uncertain. Enterprises are experimenting aggressively, but not every proof of concept becomes sustained consumption. Governments want sovereign AI capabilities, but public budgets do not move at the speed of GPU procurement. Data centers must be financed before the demand curve is fully proven.
In mature markets, hyperscalers can lean on existing enterprise bases, deep capital markets, and large ecosystems of partners. In emerging markets, the first major cloud region may need an anchor customer strong enough to convince financiers that utilization will materialize. Often, that anchor is the public sector.
That is why guaranteed capacity payments are so revealing. They are a mechanism for turning expected demand into bankable revenue. They also move risk from the provider and its partners toward the state. Whether that is sensible depends on the price, the services covered, the flexibility of the contract, the public benefit, and the transparency of the arrangement.
If the reported guarantee request was too large for Kenya, the issue may not be whether the country wants cloud infrastructure. It may be whether the proposed scale arrived ahead of demand that could be credibly contracted. Scaling back, if it happens, would be a rational response to that mismatch rather than a humiliation.
Microsoft’s Global Cloud Strategy Is Hitting Local Realities
Microsoft has been on an extraordinary infrastructure run. It has announced cloud and AI investments across Asia, Europe, Africa, and North America, often pairing data center commitments with skilling programs, government partnerships, and developer initiatives. This strategy has helped Microsoft present Azure not merely as a commercial cloud but as a platform for national competitiveness.The Kenya report shows the limits of that model. A global company can announce a standardized ambition, but it cannot standardize every country’s fiscal capacity, regulatory environment, power market, or political tolerance for guarantees. The more Microsoft frames cloud regions as national development infrastructure, the more it must deal with the messy democratic and fiscal questions that surround national infrastructure.
There is also a reputational risk. Microsoft’s cloud credibility rests partly on trust: trust that announced regions will arrive, trust that services will be durable, trust that governments and enterprises can plan around the roadmap. If a high-profile project slips, customers will ask whether other emerging-market announcements should be treated as firm commitments or strategic signaling.
That does not mean Microsoft should avoid ambitious markets. Quite the opposite. The next phase of cloud growth will depend on reaching countries and regions underserved by current infrastructure. But the company may need to be more explicit about the difference between a signed intention, a financed build, a contracted cloud region, and an operational launch.
For IT buyers, that distinction is practical. Roadmaps influence architecture. If a bank expects a local Azure region within two years, it may design modernization plans differently. If that region becomes uncertain, the bank needs a fallback strategy. Ambiguous announcements can create real planning costs.
Kenya’s Government Is Right to Keep the Door Open
Kenya’s reported position — that the project is not failed or withdrawn, but still needs structuring — is politically careful and commercially sensible. Walking away too quickly would damage a potentially valuable investment. Accepting overly burdensome guarantees could create a long-term fiscal problem and public backlash.The correct answer may be a smaller first phase, a broader base of anchor tenants, or a staged capacity model that ties expansion to actual demand. Microsoft and G42 may prefer a larger commitment because scale improves economics, but a smaller, credible deployment could be better than a grander project that cannot clear the financing and procurement hurdles.
Kenya also has leverage. Its geothermal resources, regional position, talent base, and role as an East African technology hub are real assets. Microsoft and G42 chose Kenya for reasons that have not vanished because negotiations became difficult. If anything, the continued talks suggest both sides still see value in finding a workable structure.
The government’s challenge is to avoid confusing prestige with policy. A local Azure region would be valuable, but it is not valuable at any price. The public interest depends on whether the project supports domestic capability, improves service access, strengthens resilience, and leaves Kenya with infrastructure and skills that outlast the announcement cycle.
That means transparency matters. Citizens do not need every commercial detail, but they deserve clarity about public obligations, power allocation, procurement commitments, and expected benefits. The more cloud infrastructure resembles critical national infrastructure, the less plausible it is to treat its financing as a purely private matter.
The Windows and Azure Crowd Should Read This as a Capacity Signal
For system administrators and cloud architects, the immediate operational impact is limited unless they were already planning around the proposed East Africa Azure region. Existing Azure regions remain available, and Microsoft’s global cloud is not suddenly less capable because one project is delayed. But as a signal, the story is worth attention.The most important lesson is that cloud geography remains strategic. Region selection affects latency, compliance, backup design, cost, identity architecture, and service availability. Organizations that depend on future regions should build plans that can survive schedule slips.
The second lesson is that AI demand is reshaping infrastructure priorities. Microsoft’s cloud roadmaps are increasingly tied to power availability and GPU-scale economics. That may affect not only where new regions appear, but also which services launch in those regions, how capacity is rationed, and how quickly advanced AI features become available outside the largest markets.
The third lesson is that government cloud strategy is now inseparable from energy strategy. Countries that want local AI infrastructure need reliable power, credible regulation, procurement frameworks, and enough demand aggregation to make projects financeable. Cloud adoption policy alone is no longer enough.
For enterprise customers in Kenya and neighboring markets, the prudent approach is to treat the East Africa region as promising but not guaranteed until Microsoft provides a firmer update. Design for portability where possible. Keep data residency assumptions explicit. Avoid building compliance narratives around infrastructure that is not yet operational.
The Kenya Delay Turns a Cloud Promise Into a Stress Test
The reported dispute does not prove that Microsoft’s Kenya plan is doomed. It proves that the AI infrastructure boom has entered its more serious phase, where announcements must be converted into contracts, power, capacity, and public accountability.- Microsoft and G42 announced the Kenya initiative in May 2024 as a $1 billion digital ecosystem package centered on a geothermal-powered data center and a new East Africa Azure region.
- Bloomberg News reportedly found that the project has been delayed by disagreements over Microsoft and G42’s request for guaranteed annual capacity payments from the Kenyan government.
- Kenyan officials are publicly keeping the project alive, saying it has not failed or been withdrawn while acknowledging that the original scale still needs structuring.
- The power question is as important as the payment question because AI-era cloud regions require firm, reliable, and commercially viable energy arrangements.
- A scaled-back project would not necessarily be a defeat if it produces a region that matches real demand more closely than the original proposal.
- IT leaders should treat future cloud regions as planning assumptions, not operational facts, until providers confirm timelines, services, and availability.
Source: WKZO Microsoft’s African data center falters on payment demands, Bloomberg News reports
